Scheduling and initializing software execution plans

Abstract

The invention relates to scheduling and initializing software execution plans. Embodiments of the present disclosure relate to systems and methods for scheduling and initiating execution of one or more tasks. The system may include a processing unit that may perform operations that may include obtaining execution information that may correspond to a first task and a second task. In some embodiments, a first task may include a first operation, and a second task may include a second operation. In some embodiments, the operations may also include determining a time to initialize execution of the first task and the second task based at least on the execution information. In some embodiments, execution of a first task may be performed on a first computing system, while a second task may be performed on a second computing system, where execution of the first operation and the second operation are interdependent.

Description

Background Art

[0001] Software development for certain systems (particularly complex systems, such as robotic systems, autonomous or semi-autonomous vehicles, drones, and other systems) uses a complex interaction of software and hardware to achieve a variety of results. As software becomes more advanced, the need to test various parts of the software becomes more and more pressing. In addition, as the operations performed using software become more specialized, the variability of the computing resources used to perform the specialized operations also increases.

[0002] An example challenge facing software developers is a limited number of computing resources and / or processing resources that can be configured to implement and / or execute one or more tasks and / or processes associated with software and / or related hardware. In some cases, software testing can be an important part of the development and debugging process for many complex systems. In many cases, software corresponding to multiple processes, projects, code portions, tasks, etc. may need to be tested on a limited number of computing resources, computing platforms, processors, etc.

[0003] Some traditional software testing methods may rely solely on manually performing one or more tasks or performing interdependent tasks on the same computing system. However, an example limitation of traditional methods may be the overhead and expertise required to manually perform various tasks - for example, the use time and expertise of software developers. In addition, manually performing tasks or performing tasks on the same type of computing system may reduce the efficient use of computing resources, while computing resources used in a different manner may be configured to perform more testing operations in a shorter amount of time. Summary of the invention

[0004] According to one or more embodiments of the present disclosure, a system may be configured to perform one or more operations associated with testing and / or executing one or more software workflows. For example, a language may be used to specify a workflow as a directed acyclic graph of (but not limited to) groups, where a group may include a collection of one or more tasks with parallel dependencies that must be scheduled simultaneously. The workflow specification may allow a user to build a DAG in which nodes are groups and edges represent serial dependencies between groups. In some embodiments, each task may declare specific resource requirements so that the task can be scheduled to a node with appropriate resources. By using such a workflow specification, developers (e.g., robot developers) can describe complex workflows in a reproducible manner, and can be automatically consumed and executed by a workflow execution system.

[0005] In an embodiment, the system may include a workflow engine for executing workflows submitted by users using workflow specifications. Groups may be executed in an execution backend - for example using a cluster in a container orchestration system - that supports running tasks in parallel at the same time. In some embodiments, when a user submits a workflow, the groups in the workflow are added to a database, and each group has a counter that tracks how many unfinished upstream groups it has. Groups in the workflow that have no upstream dependencies are submitted to, for example, a cluster. The system can then monitor the cluster for completed groups. When a group completes execution, the system can decrement the counters of all downstream groups of the completed group in the database. If the group's counter is decremented to zero, it means that all of its upstream groups have completed, its serial dependencies have been satisfied, and therefore it is ready to execute. At this point, the system can submit the group to the cluster for execution. This can continue until all groups in the workflow have completed.

[0006] In some embodiments, the system may include one or more processors, and the processors may be configured to perform one or more operations. In some embodiments, the operations may include obtaining execution information corresponding to the first task and the second task. In some embodiments, the first task may include one or more first containers, and the first containers may include one or more first operations, and the second task may include one or more second containers, and the second containers may include the second operation.

[0007] In some embodiments, the obtained execution information may identify one or more first dependencies that may correspond to the first task and one or more second dependencies that may correspond to the second task. In addition, the obtained execution information may include a first computing system on which the first task may be executed and a second computing system on which the second task may be executed. In some embodiments, the first computing system and the second computing system that may be identified using the obtained execution information may be different computing systems.

[0008] In some embodiments, the operation may include determining a time to initialize the execution of the first task and the second task. In some embodiments, the determined time may be based on the obtained execution information. In some embodiments, the execution of the first task on the first computing system may be initialized. Additionally or alternatively, the execution of the second task corresponding to the second computing system may be initialized. In some embodiments, the execution of the first task and the second task may be performed at the determined time. In some embodiments, the execution of one or more first containers associated with the first task and one or more second containers associated with the second task may be interdependent.

[0009] Embodiments of the present disclosure may allow for automatic execution of one or more interdependent tasks on multiple corresponding types of computing resources. In addition, embodiments of the present disclosure may improve the efficiency of using processing resources by automatically initiating the execution of interdependent tasks, which are executed simultaneously or substantially simultaneously on different types of computing resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] The present disclosure describes in detail the present system and method for scheduling and initializing software execution plans with reference to the accompanying drawings, wherein:

[0011] Figure 1 An example environment for scheduling and executing tasks corresponding to a workflow according to one or more embodiments of the present disclosure is shown;

[0012] Figure 2 is an example implementation of an execution schedule according to one or more embodiments of the present disclosure;

[0013] Figure 3A An example environment for generating one or more execution plans using one or more workflows according to one or more embodiments of the present disclosure is shown;

[0014] Figure 3B is an example implementation of a scheduler that specifies one or more containers to be executed according to one or more embodiments of the present disclosure;

[0015] Figure 4 An example environment for generating one or more outputs based on executing one or more workflows corresponding to one or more execution plans according to one or more embodiments of the present disclosure is shown;

[0016] Figure 5 is a flowchart illustrating a method for initiating execution of one or more tasks corresponding to a corresponding workflow according to one or more embodiments of the present disclosure;

[0017] Fig. 6A is an illustration of an exemplary autonomous vehicle according to some embodiments of the present disclosure;

[0018] Figure 6B According to some embodiments of the present disclosure Fig. 6A Example of camera positions and fields of view of an exemplary autonomous vehicle;

[0019] Figure 6C According to some embodiments of the present disclosure Fig. 6A A block diagram of an exemplary system architecture of an exemplary autonomous vehicle;

[0020] Fig.6DAccording to some embodiments of the present disclosure, one or more cloud-based servers and Fig. 6A A system diagram of an exemplary autonomous vehicle-to-vehicle communication system;

[0021] Figure 7 is a block diagram of an example computing device suitable for implementing some embodiments of the present disclosure; and

[0022] Figure 8 is a block diagram of an example data center suitable for implementing some embodiments of the present disclosure. DETAILED DESCRIPTION

[0023] One or more embodiments of the present disclosure may relate to software testing corresponding to one or more tasks associated with one or more workflows. "Task" may include a call to one or more software tools. In some cases, for example, each task may include several different operations that can be implemented and / or performed using one or more computing resources, for example, by activating (invoke) or calling (call) one or more functions. Executing a task may include executing one or more operations that may correspond to the task. In some cases, executing a task may include executing each operation that may be associated with the task. In some embodiments, one or more operations corresponding to a task may include a discrete executable file, which may include one or more lines of code, one or more .exe files and / or any other discrete operations that may be performed using one or more computing systems.

[0024] For example, a task may include "data processing" corresponding to a particular data set. In some cases, the task may include all operations corresponding to the data processing of the particular data set. Conversely, in other cases, the task may include a subset of operations corresponding to the data processing of the particular data set. Continuing with this example, one or more operations that may be included in the task may include one or more sub-processes implemented and / or performed when performing the data processing task. Additionally or alternatively, the operations may include one or more other discrete operations that can be performed using one or more computing systems - for example, extracting one or more portions of a data set, parsing a data set, etc.

[0025] In some cases, one or more tasks may be scheduled to run based on dependencies, available computing resources, and one or more types of computing resources (e.g., CPU, GPU, DPU, etc.) that may be indicated by one or more tasks. A collection of one or more tasks may be referred to as a "workflow" in this article. A workflow may include one or more tasks that may be interdependent. For example, a first task may include generating a test scenario for running one or more simulations corresponding to specific data. Continuing with this example, a second task may include one or more simulations that may be run using one or more generated test scenarios. Since the second task may be executed only after the first task is executed, the second task may be dependent on the first task. Further continuing with this example, in a workflow, a second task may be scheduled to be executed only after the first task can be executed.

[0026] In some embodiments, different tasks can be performed by different types of computing systems based on the indicated type of computing platform that can be used to perform the particular task. Different types of computing systems may include systems that can be used for different purposes. For example, different types of computing systems may include different types of central processing units (CPUs), graphics processing units (GPUs), tensor processing units (TPUs), field programmable gate arrays (FPGAs), etc.

[0027] According to one or more embodiments of the present disclosure, multiple tasks may be designated to be executed or otherwise scheduled to be executed as part of a workflow. Specifically, a workflow may include one or more instances, wherein a first task may be designated to be executed using a first type of computing system, and a second task may be designated to be executed simultaneously or substantially simultaneously on a second type of computing system. In addition, one or more containers corresponding to the first task and one or more containers corresponding to the second task may be interdependent. Specifically, one or more operations corresponding to a specific container of the first task may depend on one or more operations corresponding to a specific container in the second task.

[0028] One or more embodiments of the present disclosure may enable automatic execution of one or more interdependent tasks on multiple corresponding types of computing platforms. One or more embodiments described herein may improve the efficiency of using computing resources by automatically initiating execution of interdependent tasks that are executed simultaneously or substantially simultaneously on different types of computing platforms.

[0029] One or more embodiments disclosed herein may involve workflow testing and / or execution that may be performed using one or more autonomous machines, which may include any applicable machine or system capable of performing one or more autonomous and / or semi-autonomous operations. Example autonomous machines may include, but are not limited to, vehicles (land, sea, space, and / or air), robots, robotic platforms, warehouse / factory vehicles or machines, etc. For example, an autonomous machine computing application may include one or more applications that may be executed by an autonomous vehicle or a semi-autonomous vehicle, such as Figures 6A-6D An example autonomous or semi-autonomous vehicle or machine 600 (or referred to herein as "vehicle 600" or "autonomous machine 600") is depicted. In this disclosure, references to an "autonomous vehicle" or "semi-autonomous vehicle" may include any vehicle that can be configured to perform one or more autonomous or semi-autonomous navigation or driving operations. Thus, such vehicles may also include vehicles that require an operator or in which an operator may also perform such operations.

[0030] The systems and methods described herein may be used by, but are not limited to, non-autonomous vehicles or machines, semi-autonomous vehicles or machines (e.g., in one or more adaptive driver assistance systems (ADAS)), autonomous vehicles or machines, manned and unmanned robots or robotic platforms, warehouse vehicles, off-road vehicles, vehicles connected to one or more trailers, spacecraft, boats, shuttles, emergency response vehicles, motorcycles, electric or motorized bicycles, aircraft, engineering vehicles, submarines, drones, and / or other vehicle types. In addition, the systems and methods described herein may be used for a variety of purposes, such as, but not limited to, machine control, machine motion, machine driving, synthetic data generation, model training, perception, augmented reality, virtual reality, mixed reality, robotics, security and surveillance, simulation and digital twins, autonomous or semi-autonomous machine applications, deep learning, environmental simulation, object or actor simulation and / or digital twins, data center processing, conversational AI, light transport simulation (e.g., ray tracing, path tracing, etc.), collaborative content creation of 3D assets, cloud computing, and / or any other suitable application.

[0031] The disclosed embodiments may be included in a variety of different systems, such as automotive systems (e.g., control systems for autonomous or semi-autonomous machines, perception systems for autonomous or semi-autonomous machines), systems implemented using robots, aviation systems, medical systems, boating systems, smart area monitoring systems, systems for performing deep learning operations, systems for performing simulation operations, systems for performing digital twin operations, systems implemented using edge devices, systems including one or more virtual machines (VMs), systems for performing synthetic data generation operations, systems implemented at least in part in a data center, systems for performing conversational AI operations, systems implementing one or more language models (e.g., one or more large language models (LLMs) that process text, audio, images, sensors, and / or other data types to generate one or more outputs), systems for hosting real-time streaming applications, systems for presenting one or more of virtual reality content, augmented reality content, or mixed reality content, systems for performing light transport simulations, systems for performing collaborative content creation of 3D assets, systems for performing one or more generative AI operations, systems implemented at least in part using cloud computing resources, and / or other types of systems.

[0032] These and other embodiments of the present disclosure will be explained with reference to the accompanying drawings. It should be understood that these drawings are diagrammatic and schematic representations of such example embodiments and are not restrictive and are not necessarily drawn to scale. In the drawings, unless otherwise specified, features with the same number represent the same structure and function.

[0033] Reference now Figure 1 , Figure 1 An example environment 100 for scheduling and / or executing tasks corresponding to a workflow 102 according to one or more embodiments of the present disclosure is shown. In some embodiments, the environment 100 may correspond to one or more machines, systems, subsystems, and / or a collection of processes that may include software that may be tested and / or executed. For example, the workflow 102 may correspond to one or more autonomous or semi-autonomous systems, autonomous machines, robots, drones, subsystems of the aforementioned, etc.

[0034] In some embodiments, the environment 100 may include a scheduler 104, which may be configured to generate one or more workflow execution plans 106 based on the one or more workflows 102. Additionally or alternatively, the environment 100 may include an execution module 108, which may be configured to execute the one or more workflows 102 based on the one or more execution plans 106 and / or initiate execution of the one or more workflows 102.

[0035] In some embodiments, workflow 102 may include one or more tasks. In some embodiments, each task may include a call to one or more software tools. In some cases, for example, each task may include several different operations that may be implemented and / or performed using one or more computing resources, for example, by activating or calling one or more functions.

[0036] In some embodiments, operations corresponding to a particular task may be organized in one or more containers. One or more containers may include independent software packages, whereby operations organized in containers may be executed. In some instances, containers may include, for example, runtime, system tools, system settings, information on how to perform operations corresponding to containers, etc. In some embodiments, a reference to an execution container may include executing one or more operations that may be organized in a particular container. In some embodiments, a reference to an execution task may include executing one or more containers corresponding to the task. Additionally or alternatively, in some embodiments, a reference to an execution task may include executing each container corresponding to the task.

[0037] In some embodiments, the workflow 102 may include a set of tasks that can be used to achieve a specific goal. For example, the workflow 102 may include generating and / or training a specific machine learning algorithm, module, and / or system to generate a specific output. In some embodiments, training a machine learning algorithm may be performed by one or more individual tasks, for example, processing input data as a first task, training one or more artificial intelligence ("AI") models using the input data as a second task, evaluating the trained AI model as a third task, generating test scenarios as a fourth task, etc. In some embodiments, the set of tasks included in training a specific AI model may be included in the workflow 102.

[0038] In some embodiments, workflow 102 may include tasks associated with the same system, subsystem, process, etc. For example, workflow 102 may correspond to operations performed using a perception system or subsystem corresponding to a self-machine. Continuing with this example, workflow 102 may include a first task that may be configured to train a machine learning model to recognize an object based on input data (e.g., image data, radio detection and ranging (“RADAR”) data, light detection and ranging (“LiDAR”) data, etc.). Continuing with this example, workflow 102 may also include a second task that may correspond to generating one or more control commands based on data that may be generated using a machine learning model that may have been trained using the first task.

[0039] In contrast, in some embodiments, the workflow 102 may include tasks that may not be associated with the same system, subsystem, process, etc. Instead, the tasks included in the respective workflow 102 may include tasks corresponding to separate systems, subsystems, processes, etc. For example, the workflow 102 may include a first task that may correspond to a perception system or subsystem that may be associated with the self-machine. Continuing with the example, the workflow 102 may include a first task that may be configured to train one or more machine learning models to recognize objects based on input data (e.g., image data, RADAR data, LiDAR data, etc.). Further continuing with the example, the workflow 102 may include a second task that may correspond to a communication subsystem associated with the self-machine. The second task may be configured to detect one or more errors corresponding to one or more communication systems that may be configured to send and / or receive data.

[0040] In some embodiments, an individual workflow 102 may include one or more tasks that may be dependent on one or more other tasks within the individual workflow 102. For example, a workflow 102 may include a first task and a second task. The first task may include generating a test scenario for running one or more simulations corresponding to specific data. Continuing with the example, the second task may include one or more simulations that may be run using one or more generated test scenarios. Since the second task can only be executed after the first task is executed, the second task may be dependent on the first task. Further continuing with the example, in the workflow 102, the second task may be scheduled to be executed only after the first task is executed.

[0041] In some embodiments, two or more tasks corresponding to the workflow 102 may be interdependent. Two or more interdependent tasks included in the workflow 102 may be referred to herein as a "group". For example, the workflow 102 may include a group that includes a first task and a second task. The first task may include one or more operations that may be dependent on the execution of the second task, and correspondingly, the second task may include one or more operations that may be dependent on the execution of the first task. Further continuing with this example, because the first task and the second task may include interdependent operations, the first task and the second task may be executed simultaneously or substantially simultaneously.

[0042] In some embodiments, each task may be designated for execution by a different computing platform. In some embodiments, the computing platform may include one or more environments, including hardware and software on which applications and software systems run and / or operate. Different types of computing platforms may include a central processing unit (CPU), a graphics processing unit (GPU), a tensor processing unit (TPU), a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), an accelerator (e.g., a deep learning accelerator (DLA)), and / or other computing platforms. Additionally or alternatively, different types of computing platforms may include different types of CPUs, GPUs, TPUs, FPGAs, ASICs, DLAs, accelerators, etc. For example, a first type of GPU may be more suitable for image and / or video processing, while a second type of GPU may be more suitable for training a machine learning model. Continuing with this example, a first type of GPU may be a first type of computing platform, while a second type of GPU may be a second type of computing platform.

[0043] In some embodiments, a specific task may be designated to be performed only by a specific computing platform. For example, a first task may be designated to be performed only using a specific GPU. Additionally or alternatively, a specific task may be designated to be performed by multiple different computing platforms. For example, a first task may be designated to be performed using multiple different types of CPUs and multiple different types of GPUs.

[0044] In some embodiments, workflow 102 may specify that one or more tasks may be performed using different computing platforms. In some embodiments, a first task may be specified to be performed using a first type of computing platform, and a second task may be specified to be performed using a second type of computing platform. For example, workflow 102 may specify a first task and a second task for execution, wherein the first task and the second task may be associated with the same system or machine (e.g., a self-machine). Continuing with this example, the first task may include one or more operations related to signal processing that may be performed using one or more FPGAs. Further continuing with this example, the second task may include one or more image processing operations that may be performed using one or more GPUs.

[0045] In some embodiments, the workflow 102 may be generated using details provided by one or more systems, entities, users, etc. For example, a system may generate a workflow (e.g., a specification or directed acyclic graph ("DAG")) associated with executing a set of tasks that may be associated with the workflow 102. In some embodiments, the workflow 102 may include one or more interdependencies between one or more tasks and / or containers included in the one or more tasks. Additionally or alternatively, one or more users, entities, systems, etc. may specify a type of computing platform on which one or more tasks associated with the workflow 102 and / or containers corresponding to the one or more tasks may be executed.

[0046] In some embodiments, the workflow 102 may include execution information corresponding to tasks and / or containers corresponding to the workflow 102. In some embodiments, the execution information corresponding to the workflow 102 may specify tasks to be executed and / or containers corresponding to the specified tasks. In some embodiments, the execution information may include information about a computing platform for executing one or more tasks and / or containers associated with the workflow 102. In some embodiments, the workflow 102 may include information about dependencies between tasks, containers, and / or operations associated with the workflow 102.

[0047] Additionally or alternatively, the execution information corresponding to the workflow 102 may include timeout information indicating a time threshold that may indicate when the task is automatically ended in response to the task not being completed. Additionally or alternatively, the execution information may indicate a system, user, entity, etc. that may be requesting to execute the task corresponding to the workflow 102. In some embodiments, the execution information corresponding to the workflow may indicate an amount of computing resources required to execute the task corresponding to the workflow 102.

[0048] In some embodiments, the workflow 102 may be sent, transmitted, or otherwise communicated to the scheduler 104. In some embodiments, the workflow 102 may be received by the scheduler 104 and / or otherwise obtained by the scheduler 104. In some embodiments, the scheduler 104 may be configured to schedule one or more tasks corresponding to the workflow 102 for execution based on the workflow 102 and the information included therein.

[0049] In some embodiments, the scheduler 104 may include code and routines configured to allow a computing system to perform one or more operations. Additionally or alternatively, the scheduler 104 may be implemented using hardware including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., for executing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), and / or other processor types. In these and other embodiments, the scheduler 104 may be implemented using a combination of hardware and software. In the present disclosure, the operations performed by the scheduler 104 may include operations that the scheduler 104 may instruct the corresponding computing system to perform. In these or other embodiments, the scheduler 104 may be implemented by one or more computing devices, such as with respect to Figures 6A-6D , Figure 7 and / or Figure 8 A computing device is described in more detail.

[0050] In some embodiments, the scheduler 104 may be configured to generate one or more workflow execution plans 106 based on the execution information provided in one or more workflows 102. For example, the scheduler 104 may be configured to schedule tasks corresponding to the workflow 102 based on dependencies corresponding to the tasks. For example, where a second task depends on a first task, the second task is scheduled to be executed after the first task. In some embodiments, the scheduler 104 may indicate the execution order and / or timing of the tasks. In these or other embodiments, the scheduler 104 may instruct different computing platforms (e.g., GPUs, CPUs, accelerators, VPUs, DPUs, PPUs, etc.) to manage the execution of corresponding tasks.

[0051] In some embodiments, the scheduler 104 may be configured to schedule one or more tasks for execution based on the amount of computing resources available to execute the one or more tasks. For example, various tasks that may be executed on the same computing platform may be scheduled to be executed sequentially or simultaneously based on the computing resources available to execute the various tasks. In some embodiments, the scheduler 104 may be configured to schedule one or more tasks corresponding to the workflow 102 for execution based on the system, user, entity, etc. that may be requesting the execution of the task. For example, certain systems, entities, users, etc. may be given a priority status compared to one or more other tasks and / or workflows 102.

[0052] In some embodiments, the scheduler 104 may be configured to schedule one or more tasks for execution based on a computing platform designation. For example, one or more workflows 102 may specify one or more computing platforms on which one or more tasks may be executed. In some embodiments, the scheduler 104 may be configured to schedule one or more tasks for execution based on corresponding designations corresponding to one or more tasks included in the workflow 102.

[0053] In some embodiments, the scheduler 104 may be configured to schedule one or more tasks to be executed based on one or more groups that may correspond to the identification of the workflow 102. For example, a first group may include a first task and a second task that may be interdependent, so the first task and the second task may be executed simultaneously or substantially simultaneously. In some embodiments, the scheduler 104 may only schedule the first task and the second task to be executed when computing resources are available to process both the first task and the second task simultaneously or substantially simultaneously.

[0054] In some embodiments, the scheduler 104 may be configured to dynamically schedule one or more tasks. In some embodiments, the scheduler 104 may be configured to generate a task queue based on one or more workflows 102 for execution. In some embodiments, the scheduler 104 may be configured to specify one or more tasks in the generated queue for execution at runtime. In some embodiments, the designation of one or more tasks may be determined based on the execution information corresponding to one or more workflows 102 and the availability of computing resources during runtime. In some embodiments, dynamically scheduling one or more tasks in real time or substantially real time can improve the efficient utilization of computing resources and / or reduce the amount of time spent performing tasks corresponding to one or more workflows 102.

[0055] In some embodiments, the availability of computing resources may be determined and / or evaluated at runtime. In some embodiments, the current availability of computing resources corresponding to one or more computing platforms may be evaluated to dynamically schedule tasks for execution.

[0056] For example, the first workflow 102 may include a first task, and the second workflow 102 may include a second task, wherein both the first task and the second task may be queued to be scheduled for execution. Continuing with this example, the first task may require a first amount of computing resources to execute, and the second task may require a second amount of computing resources to execute, wherein the first amount of computing resources is greater than the second amount of computing resources. Further continuing with this example, both the first task and the second task may be designated as being executed on the same computing platform. Before determining whether to schedule one or more of the first task and the second task, the current availability of the computing resources corresponding to the computing platform may be evaluated. In some cases, the scheduler 104 may determine that only enough computing resources are available for scheduling the second task for execution. Additionally or alternatively, in some cases, the current availability of computing resources may be sufficient to schedule both the first task and the second task together. Additionally or alternatively, in some cases, the current availability of computing resources may not be sufficient to schedule the first task or the second task, and the scheduler 104 may be configured to wait until computing resources become available for scheduling the first task and / or the second task for execution.

[0057] In some embodiments, the scheduler 104 may generate one or more execution plans 106, which may indicate an execution order and / or timing of tasks. For example, the one or more execution plans 106 may indicate an execution order of tasks, which computing platform may execute the corresponding tasks, and / or timing constraints on the tasks (e.g., how much time is allocated for each corresponding task and / or timings on when the corresponding tasks should start and / or when they should end).

[0058] In some embodiments, one or more execution plans 106 may include one or more stages in which one or more tasks may be scheduled to be executed. In some embodiments, a stage may refer to a discrete period in which one or more tasks may be executed. In some embodiments, one or more tasks corresponding to a workflow 102 may be executed at different stages based on dependencies between tasks, the amount of available computing resources, the amount of time to execute tasks, and / or other factors (these factors may indicate that a task may be executed at a specific stage, which may be different or the same as the stage in which one or more other tasks may be executed). In some embodiments, a group may be scheduled in a single stage so that two or more tasks included in the group may be executed simultaneously or substantially simultaneously.

[0059] In some embodiments, different phases may include tasks that may be performed one after another at discrete times. For example, in a workflow including two phases, a first phase may be performed, and after the first phase is completed, a second phase may begin. Additionally or alternatively, the phases may be performed simultaneously. For example, in a workflow including two phases, one or more tasks corresponding to the first phase may begin to be performed, and before one or more tasks complete execution, one or more tasks corresponding to the second phase may begin to be performed.

[0060] In some embodiments, an execution plan 106 indicating that one or more tasks may be performed at a particular stage may indicate that execution of the stage may be initiated, and then execution of a subsequent stage may be initiated. For example, a workflow 102 may include a first stage and a second stage, wherein the first stage may include a first task to be performed, and the second stage may include a second task to be performed. Continuing with this example, the execution plan 106 may indicate that a first task corresponding to the first stage may be performed first, and that a second task corresponding to the second stage may be performed at a certain period of time after the start of execution of the first task corresponding to the first stage.

[0061] In some embodiments, one or more execution plans 106 may be transmitted, sent, or otherwise communicated to execution module 108. In some embodiments, execution module 108 may be configured to receive and / or otherwise obtain one or more execution plans 106.

[0062] In some embodiments, the execution module 108 may include code and routines configured to allow the computing system to perform one or more operations. Additionally or alternatively, the execution module 108 may be implemented using hardware including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., for executing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), and / or other processor types. In these and other embodiments, the execution module 108 may be implemented using a combination of hardware and software. In the present disclosure, operations described as being performed by the execution module 108 may include operations that the execution module 108 may instruct the corresponding computing system to perform. In these or other embodiments, the execution module 108 may be implemented by one or more computing devices, such as with respect to Figures 6A-6D , Figure 7 and / or Figure 8 A computing device is described in further detail.

[0063] The execution module 108 may be configured to execute tasks corresponding to the workflow 102 that are scheduled for execution according to the execution plan 106. In some embodiments, the execution module 108 may be configured to initiate execution of one or more tasks corresponding to the corresponding workflow 102 based on one or more execution plans 106, rather than the execution module 108 executing the tasks corresponding to the corresponding workflow 102, wherein the execution is performed using one or more other computing resources (e.g., one or more CPUs, GPUs, DPUs, FPGAs, etc.).

[0064] In some embodiments, the execution module 108 may be configured to monitor one or more tasks being executed. In some embodiments, the execution module 108 may be configured to receive and / or extract one or more errors that may have been generated during the execution of the tasks. For example, the errors may include any number of errors, such as syntax errors corresponding to the code, type errors, resource errors, network errors, errors corresponding to hardware limitations, memory errors, and any other errors that may be encountered during the execution of the tasks corresponding to the corresponding workflow 102.

[0065] In some embodiments, the execution module 108 can be configured to pause and / or stop the execution of one or more tasks by one or more computing platforms. In some embodiments, the execution module 108 can stop and / or pause the execution of one or more tasks based on one or more detected errors. Additionally or alternatively, the execution module 108 can be configured to stop and / or pause the execution of one or more tasks based on each task being executed for more than a predetermined amount of time. In some embodiments, the predetermined amount of time can be a multiple of the expected amount of time to perform the task. In some embodiments, the expected amount of time can be included in the workflow 102, and can be included to efficiently use computing resources when a task can fall into a loop and run forever.

[0066] In some embodiments, the execution module 108 can be configured to generate a queue of tasks that can be run based on the amount of computing resources currently available. For example, the workflow execution plan 106 can indicate that multiple different tasks can be executed using the first computing platform. Continuing with this example, in the case where the CPU cannot handle the execution of multiple tasks at once, the execution module 108 can manage the queue based on one or more workflow execution plans 106 associated with the tasks.

[0067] In some embodiments, the execution module 108 may be additionally configured to extract execution information generated based on executing one or more tasks. In some embodiments, for example, a log may be generated based on executing one or more tasks, and the execution module 108 may be configured to send and / or transmit the log to one or more other systems, machines, users, entities, etc.

[0068] In some embodiments, execution module 108 may be configured to generate output 110. In some embodiments, output 110 may include information corresponding to the execution of one or more tasks.

[0069] In some embodiments, output 110 may include information corresponding to detected errors associated with the execution of one or more tasks. Additionally or alternatively, output 110 may include a notification including, for example, that execution module 108 stops and / or pauses the execution of one or more tasks.

[0070] In some embodiments, the output 110 may be sent or otherwise communicated to one or more other systems, entities, users, etc. For example, the output 110 may be communicated to an entity or system that may have generated one or more workflows 102 .

[0071] Can Figure 1 Modifications, additions, or omissions may be made without departing from the scope of the present disclosure. For example, the number of tasks included in each workflow 102, the number of workflows 102, and the number of execution plans 106 may vary. The details given and discussed are intended to help explain and understand the concepts of the present disclosure and are not intended to be limiting.

[0072] Figure 2 is an example implementation of an execution plan 200 according to one or more embodiments of the present disclosure. In these and other embodiments, the execution plan 200 may be an example of an implementation of an execution plan 200 according to one or more embodiments of the present disclosure. Figure 1 Examples of one or more execution plans 106 and / or the like are further described and / or illustrated. In some embodiments, the execution plan 200 may be an example of scheduling tasks corresponding to a particular workflow 102, which may be described in the present disclosure, for example, with respect to Figure 1 Additionally or alternatively, the execution plan 200 may be an example of scheduling tasks corresponding to a plurality of workflows 102 that may be described in the present disclosure, for example, with respect to Figure 1 Further description and / or explanation is provided.

[0073] In some embodiments, execution plan 200 may specify that one or more tasks 204 may be executed. In some embodiments, execution plan 200 may specify first task 204A, second task 204B, third task 204C, fourth task 204D, fifth task 204E, sixth task 204F, and seventh task 204G, collectively referred to as "tasks 204" to be executed. In some embodiments, each corresponding task may include one or more corresponding operations. In addition, references to executing task 204 may include performing one or more operations corresponding to task 204. In some embodiments, task 204 may be associated with, for example, a task described in this disclosure with respect to Figure 1 The tasks described in one or more of the portions of workflow 102 described and / or illustrated further are the same and / or similar.

[0074] In some embodiments, task 204 may be based on one or more dependencies, current availability of computing resources, time required to perform task 204, and other aspects of the disclosure, such as those related to Figure 1 Scheduled based on other considerations described further.

[0075] In some embodiments, tasks 204 may be assigned and / or allocated to one or more phases 206, each of which may represent a period of time during which one or more tasks 204 may be performed. Figure 2 The execution plan 200 shown may include a first stage 206A, a second stage 206B, a third stage 206C, a fourth stage 206D, and a fifth stage 206E (collectively referred to as "stages 206"). In some embodiments, each stage 206 may include one or more tasks 204 to be performed. For example, the first stage 206A may include a first task 204A to be performed, the second stage 206B may include a second task 204B to be performed, the third stage 206C may include a third task 204C to be performed, and so on.

[0076] In some embodiments, the various stages 206 may be performed sequentially. For example, the first task 204A corresponding to the first stage 206A may be performed, and then the second task 204B corresponding to the second stage 206B may be performed, and so on. In some embodiments, the various stages may be performed separately. In some embodiments, rather than performing the tasks 204 corresponding to the corresponding stages 206 one after the other, the tasks 204 corresponding to the corresponding stages 206 may be performed at any time after any dependencies are satisfied. For example, the first task 204A corresponding to the first stage 206A may be performed. Continuing with the example, the second task 204B corresponding to the second stage 206B may then be performed at a later time period.

[0077] In some embodiments, each stage 206 may be executed to encourage efficient use of computing resources. For example, a first task 204A corresponding to the first stage 206A may be executed. Thereafter, another first task corresponding to another first stage associated with another execution plan may be executed. Continuing with this example, after completing execution of another first task corresponding to another execution plan, a second task 204B corresponding to the second stage 206B associated with the execution plan 200 may then be executed.

[0078] In some embodiments, multiple tasks 204 may be designated for execution in a single stage 206. In some embodiments, one or more groups (e.g., group 208) may include two or more tasks 204 that may be interdependent. In some embodiments, two or more interdependent tasks 204 may each include operations whose execution may be interdependent. As shown in execution plan 200, fifth stage 206E may include group 208.

[0079] In some embodiments, the group 208 may include a fifth task 204E, a sixth task 204F, and a seventh task 204G. In some embodiments, the fifth task 204E, the sixth task 204F, and the seventh task 204G may include operations whose execution may be interdependent. In some embodiments, in response to the tasks 204 included in the group 208 being interdependent, the fifth task 204E, the sixth task 204F, and the seventh task 204G may be designated to be executed simultaneously or substantially simultaneously.

[0080] In some embodiments, the execution plan 200 may indicate one or more computing platforms 202 on which one or more tasks 204 corresponding to the corresponding stages 206 may be executed. In some embodiments, the execution plan 200 may include a first computing platform 202A, a second computing platform 202B, and a third computing platform 202C. In some embodiments, the first computing platform 202A may be different from the second computing platform 202B, and the second computing platform 202B may be different from the third computing platform 202C. For example, the first computing platform 202A may include a CPU, the second computing platform 202B may include a GPU, and the third computing platform may include a second type of GPU.

[0081] In the execution plan 200, the columns corresponding to the computing platforms 202 may indicate which tasks 204 may be designated for execution using each computing platform 202. For example, the third task 204C and the sixth task 204F may be designated for execution using the first computing platform 202A. Continuing with this example, the first task 204A, the fourth task 204D, the fifth task 204E, and the seventh task 204G may be designated for execution using the second computing platform 202B. In addition, the second task 204B may be designated for execution using the third computing platform 202C.

[0082] In some embodiments, computing platforms 202 may execute different tasks 204 simultaneously or substantially simultaneously to satisfy interdependencies between tasks 204 included in one or more groups 208. For example, during fifth stage 206E, sixth task 204F may be designated to be executed on first computing platform 202A simultaneously or substantially simultaneously with fifth task 204E and seventh task 204G being executed on second computing platform 202B.

[0083] In these and other embodiments, a computing platform 202 that can perform one or more tasks 204 can, for example, use the present disclosure, for example, with respect to Figure 1 One or more workflows 102 further described are identified or otherwise determined.

[0084] Figure 2 Also shown is an example end-to-end workflow that may include operations corresponding to various tasks 204. In some embodiments, the execution of various tasks 204 may use one or more schedulers (e.g., Figure 1 Additionally or alternatively, the execution of the various tasks 204 may be initiated using one or more execution modules (eg, execution module 108) further described and / or illustrated in this disclosure.

[0085] In some embodiments, workflow 200 may be based on one or more DAGs, specifications, execution instructions, etc., which may indicate one or more tasks 204 that may be performed. In the example shown, workflow 200 may include a first stage 206A, a second stage 206B, a third stage 206C, a fourth stage 206D, and a fifth stage 206E, wherein one or more tasks 204 may be scheduled for execution. In some embodiments, workflow 200 may result in training one or more machine learning models, and testing software corresponding to one or more machines based on the one or more trained machine learning models.

[0086] In some embodiments, workflow 200 may include: a first computing platform 202A, which may include a first type of CPU; a second computing platform 202B, which may include multiple CPUs and GPUs that may be configured for rendering; and a third computing platform 202C, which may include multiple GPUs that may be specifically designed for machine learning. In some embodiments, each of first computing platform 202A, second computing platform 202B, and third computing platform 202C may be different from each other, so that different tasks 204 may be scheduled on first computing platform 202A, second computing platform 202B, or third computing platform 202C based on execution instructions corresponding to tasks 204.

[0087] The first task 204A may include data processing, the second task 204B may include training an AI model, the third task 204C may include evaluating the accuracy of the trained AI model from the second task 204B, the fourth task 204D may include generating one or more test scenarios corresponding to the AI ​​model and / or hardware corresponding to the AI ​​model, the fifth task 204E may include simulating a test scenario corresponding to the fourth task 204D, the sixth task 204F may include testing software during simulation of the fifth task 204E, and the seventh task 204G may include visually debugging the software being tested corresponding to the sixth task 204F. The workflow 200 including the tasks 204 may be described in further detail below.

[0088] For example, the first task 204A may include data processing. In some embodiments, data processing may include performing one or more operations on one or more data sets. In some embodiments, data processing may include data collection, data maintenance (e.g., data standardization, deduplication, data correction, etc.), data conversion, generating synthetic data, etc. In some embodiments, data processing corresponding to the first task 204A may be performed when preparing to perform one or more additional tasks (e.g., the second task 204B, the third task 204C, etc.). For example, data processing corresponding to the first task 204A may include generating synthetic data for use, for example, in the second task 204B, and the second task 204B may include training a specific AI model. In some embodiments, data processing corresponding to the first task 204A may be performed using a second computing platform 202B (multiple CPUs and GPUs specifically designed and / or configured for rendering).

[0089] In some embodiments, data processing corresponding to the first task 204A may be performed during the first phase 206A. In some embodiments, the first phase 206A may be completely separate from the second phase 206B. Additionally or alternatively, the second phase 206B may begin executing the second task at some point after the first task 204A has begun execution. In some embodiments, the tasks 204 performed during the second phase 206B, the third phase 206C, etc. may be performed concurrently with the first phase 206A and / or a portion of the first phase 206A.

[0090] In some embodiments, the second task 204B may include training one or more AI models using data obtained, for example, from the data processing of the first task 204A. In some embodiments, the AI ​​training may be performed using a third computing platform 202C including a plurality of GPUs specifically designed and configured for machine learning. In some embodiments, training the AI ​​model corresponding to the second task 204B may include training one or more models to perform a specific task. In some cases, the specific task may correspond to one or more of the tasks 204 associated with the workflow 200.

[0091] In some embodiments, the second task 204B may be performed during the second stage 206B. The second stage 206B may be performed after and / or concurrently with the first task 204A. In some embodiments, the second stage 206B may be performed before and / or concurrently with the tasks corresponding to the third stage 206C, and so on.

[0092] The third task 204C may include evaluating one or more aspects of the AI ​​model that has been trained during the execution of the second task. For example, the accuracy corresponding to the AI ​​model may be evaluated based on one or more criteria. For example, in some cases, the accuracy of the AI ​​model may be evaluated based on the target hardware on which the trained AI model may be used. In some embodiments, the third task 204C may be performed on the first computing platform 202A and during the third stage 206C.

[0093] In some embodiments, one or more test scenarios corresponding to the fourth task 204D may be generated. The test scenario may be generated based on an AI model that may have been trained and / or evaluated and associated with the second task 204B and / or the third task 204C. In some embodiments, one or more test scenarios may be generated using the second computing platform 202B. In addition, in some embodiments, one or more generated test scenarios may be used to perform one or more additional tasks (e.g., the fifth task 204E, the sixth task 204F, and / or the seventh task 204G).

[0094] In some embodiments, the fifth task 204E, the sixth task 204F, and the seventh task 204G may be initialized to be executed simultaneously as a group (e.g., group 208). In some embodiments, each task 204 corresponding to the group 208 may be executed simultaneously. Additionally or alternatively, the tasks 204 corresponding to the group may include some overlapping execution times. In some embodiments, the fifth task 204E, the sixth task 204F, and the seventh task 204G may include one or more operations that may be dependent on each other, such that the execution of one of the tasks 204 is dependent on one of the other tasks 204 corresponding to the group 208.

[0095] In this example, the fifth task 204E may include simulating one or more trained and / or evaluated AI models using a test scenario that may have been generated corresponding to the fourth task 204D. In some embodiments, the sixth task 204F may include testing software corresponding to the simulation that may be performed corresponding to the fifth task 204E. In some embodiments, the software may be tested together with the trained AI model using, for example, one or more test scenarios. In addition, the seventh task may include visually debugging software associated with the software under test corresponding to the sixth task 204F. In some embodiments, the simulation corresponding to the fifth task 204E may be performed simultaneously with the software test corresponding to the sixth task 204F, and simultaneously with the visual debugging of the software under test corresponding to the seventh task 204G.

[0096] In some embodiments, group 208 can be executed simultaneously or substantially simultaneously. In addition, in some embodiments, the software under test can be executed using a CPU corresponding to first computing platform 202A, and simulation and visual debugging can be performed using multiple CPUs and GPUs associated with second computing platform 202B.

[0097] In some embodiments, one or more outputs may be generated by executing the tasks 204 corresponding to the group 208. For example, a log corresponding to testing the software may be generated, results of one or more simulations corresponding to the fifth task 204E may be generated, etc. In some embodiments, the one or more outputs may be transmitted or otherwise communicated to one or more systems, entities, machines, etc.

[0098] Without departing from the scope of this disclosure, Figure 2200. The present invention also provides a method for implementing a plurality of execution plans ...

[0099] Figure 3A An example environment 300 for generating one or more execution plans 306 using one or more workflows 302 according to one or more embodiments of the present disclosure is shown. In some embodiments, the environment 300 may include a scheduler 304, which is included in a runtime system 314, and the scheduler 304 may be configured to perform one or more operations using information corresponding to the workflow 302 to generate the one or more execution plans 306.

[0100] One or more workflows 302 may include a set of tasks that can be performed. In some embodiments, one or more workflows 302 may be associated with the present disclosure, for example, with respect to Figure 1 The workflow 102 described and / or illustrated further is the same and / or similar.

[0101] In some embodiments, the information associated with the workflow 302 may include an estimated time period during which the corresponding task can be executed. For example, it may be estimated that the first task can be executed within a certain number of hours—for example, 6 hours, 12 hours, 24 hours, etc. In addition, in some embodiments, the information associated with one or more workflows 302 may include one or more timeout periods corresponding to the corresponding task. The timeout period may indicate the amount of time that the specific task is kept running before the execution of the specific task is canceled and / or suspended. In some embodiments, for example, a specific timeout period may be a multiple of the estimated execution time corresponding to an individual task. For example, a specific task may correspond to an estimated execution time. Continuing with this example, a specific task may also correspond to a timeout period of 2 times the estimated execution time, a timeout period of 3 times the estimated execution time, and the like.

[0102] Additionally or alternatively, information associated with one or more workflows 302 may include a network topology on which tasks, containers, and / or corresponding operations may be executed. The network topology includes, for example, a layout of network connections, nodes, links, devices, etc. corresponding to a computing platform. For example, one or more tasks, containers, and / or operations may be interdependent, and thus, the interdependent tasks, containers, and / or operations may be scheduled on computing platforms having the same or similar network topology (e.g., good network connections between computing platforms).

[0103] Furthermore, the information corresponding to the workflow 302 may include metadata, such as a timestamp indicating when the corresponding workflow 302 was generated and / or otherwise created. The metadata may additionally include user and / or system information, including information indicating a system, subsystem, user, developer, entity, etc., to which the workflow 302 may correspond.

[0104] In some embodiments, the scheduler 304 can be configured to generate one or more execution plans 306. In these and other embodiments, the scheduler 304 can be configured to generate one or more execution plans 306. Figure 1 In some embodiments, the scheduler 304 may be included in the runtime system 314 so that one or more execution plans 306 may be generated in real time or substantially in real time, similar to the scheduler 106 described and / or illustrated in the further description. Figure 1 The execution plans 106 generated by the scheduler 106 described above are very similar.

[0105] In some embodiments, scheduler 304 may correspond to one or more scheduling platforms (e.g., one or more containerized scheduling platforms) that may be configured to schedule one or more tasks corresponding to workflow 302 for execution. Some example implementations of containerized scheduling and / or orchestration platforms include Kubernetes, Docker Swarm, Amazon Elastic Container Serve ("ECS"), Google Cloud Run, and / or other scheduling platforms that may schedule and / or otherwise specify one or more tasks corresponding to one or more workflows 302 to be executed.

[0106] In some embodiments, the scheduler 304 may additionally include one or more modules that may be configured to perform one or more operations using information corresponding to the workflow 302. The one or more modules may provide additional functionality for scheduling tasks corresponding to one or more workflows 302. In some embodiments, the scheduler 304 may include a co-scheduling module 308, a colocation module 310, and / or a priority escalation module 312. In some embodiments, the various modules included in the scheduler 304 may facilitate the generation and / or creation of the execution plan 306.

[0107] In some embodiments, the co-scheduling module 308, the hosting module 310 and / or the priority upgrade module 312 may include code and routines configured to allow the computing system to perform one or more operations. Additionally or alternatively, the co-scheduling module 308, the hosting module 310 and / or the priority upgrade module 312 may be implemented using hardware, which includes one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., for executing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), and / or other processor types. In these and other embodiments, the co-scheduling module 308, the hosting module 310 and / or the priority upgrade module 312 may be implemented using a combination of hardware and software. In the present disclosure, operations described as being performed by any corresponding module may include operations that the module may instruct the corresponding computing system to perform. In these or other embodiments, the execution module 108 may be implemented by one or more computing devices, such as with respect to Figures 6A-6D , Figure 7 and / or Figure 8 A computing device is described in further detail.

[0108] The co-scheduling module 308 may be configured to receive and / or otherwise obtain the workflow 302 and / or information corresponding to the workflow 302. In some embodiments, the co-scheduling module 308 may be configured to schedule operations corresponding to one or more tasks in the one or more execution plans 306 based on one or more workflows 302 and / or information corresponding to the one or more workflows 302. In some embodiments, the co-scheduling module 308 may be configured to schedule operations corresponding to specific tasks together. In some embodiments, scheduling operations corresponding to specific tasks together may help improve the efficiency and availability of computing resources corresponding to a computing platform that may execute one or more tasks.

[0109] In some embodiments, the co-scheduling module 308 may be configured to dynamically schedule one or more tasks and / or containers corresponding to the corresponding tasks. In some embodiments, the co-scheduling module 308 may be configured to generate a queue, which may include one or more tasks and / or corresponding containers associated with one or more workflows 302. In some embodiments, the queue of tasks and / or corresponding containers may include multiple queues corresponding to tasks and / or corresponding containers that may be designated to be executed using a specific computing platform. For example, a first queue may correspond to a first set of tasks and / or containers designated to be executed on a first computing platform. Continuing with this example, a second queue may correspond to a second set of tasks and / or containers that may be designated to be executed on a second computing platform.

[0110] In some embodiments, one or more queues may indicate an order in which corresponding tasks and / or corresponding containers may be executed. The order may be determined based on one or more workflows 302 and / or information corresponding to one or more workflows 302. In some embodiments, the co-scheduling module 308 may be configured to dynamically schedule one or more tasks and / or containers based on one or more workflows, information corresponding to one or more workflows 302, and / or current availability of computing resources corresponding to various computing platforms.

[0111] For example, the collaborative scheduling module 308 may receive a first workflow 302 and a second workflow 302. The first workflow 302 may include a first task, the first task including a first container, a second container, and a third container, wherein the first container, the second container, and the third container may all be scheduled to execute on a first computing platform. Continuing with the example, the second workflow 302 may include a first task, the first task may include a first container, wherein the first container may be scheduled to execute on a first computing platform. Further continuing with the example, the first computing platform may include computing resources for executing two containers simultaneously. The collaborative scheduling module 308 may be configured to schedule the first container and the second container associated with the first task to be executed on the first computing resource. Thereafter, the collaborative scheduling module 308 may be configured to schedule the first container corresponding to the first task of the second workflow 302 to be scheduled together with the third container corresponding to the first task of the first workflow 302.

[0112] In some embodiments, by scheduling one or more containers corresponding to multiple workflows to be executed based on currently available computing resources, computing resources can be used more efficiently than if containers corresponding to individual tasks were separately scheduled and / or designated for execution. In contrast, the collaborative scheduling module 308 can be configured to schedule and / or designate only containers corresponding to specific tasks for execution. For example, each container corresponding to a first task is scheduled, and then each container corresponding to a second task is scheduled.

[0113] In some embodiments, the co-scheduling module 308 may be configured to schedule a container group for execution. In some embodiments, multiple containers may be designated to be executed together. In some embodiments, the co-scheduling module 308 may schedule the container group for execution only when computing resources become available to execute each container included in the container group simultaneously or substantially simultaneously. In some embodiments, in the case where computing resources are not yet available, the co-scheduling module 308 may schedule one or more other containers for execution based on the availability of computing resources and information corresponding to one or more workflows 302.

[0114] For example, the workflow 302 may include a first task, which may include a first set of operations and a second set of operations, wherein the first set of operations may be organized in a first container and the second set of operations may be organized in a second container. Continuing with this example, the workflow 302 may specify a specific computing platform on which the first task may be executed. In addition, the workflow 302 may specify that the first set of operations (e.g., the first container) and the second set of operations (e.g., the second container) may be executed simultaneously or substantially simultaneously. In this example, the collaborative scheduling module 308 may generate an execution plan 306, which may queue the first set of operations and the second set of operations to be executed together simultaneously or substantially simultaneously.

[0115] As a comparative example, a workflow 302 may include a first task that may include a first set of operations and a second set of operations, wherein the first set of operations may be organized in a first container and the second set of operations may be organized in a second container. Continuing with this example, the workflow 302 may specify a specific computing platform on which the first task may be specified to be executed. In addition, the workflow 302 may not include a designation that the first set of operations (e.g., the first container) and the second set of operations (e.g., the second container) may be executed simultaneously or substantially simultaneously. In this example, the co-scheduling module may generate an execution plan 306 that may specify that the first set of operations may be executed at a different time or at the same time as the second set of operations.

[0116] An example implementation of the co-scheduling module 308 in combination with the scheduler 304 may be Figure 3B described and / or shown in Figure 3B An example environment 350 is shown, which includes an implementation for specifying one or more containers associated with respective tasks for execution, wherein the respective tasks may correspond to one or more workflows—for example, in the present disclosure, for example, respectively, with respect to Figure 1 and workflow 102 and / or workflow 302 as further described and / or illustrated in FIG. 3 .

[0117] refer to Figure 3B , Figure 3B is an example implementation 350 in which a scheduler 332 designates one or more containers to be executed in accordance with one or more embodiments of the present disclosure. In some embodiments, the example implementation 350 may include a first task 320 and a second task 324, wherein the first task 320 and the second task 324 may be designated for execution.

[0118] In some embodiments, the first task 320 may include a first container 322A and a second container 322B. In some embodiments, the first container 322A may include a first set of operations, and the second container may include a second set of operations 322B. In some embodiments, the first container 322A and the second container 322B may be as described in the present disclosure, for example, with respect to Figure 1 , Figure 2 and Figure 3A Examples of containers are further described and / or illustrated.

[0119] In some embodiments, the first task 320 may correspond to a workflow, such as, for example, Figure 1 and Figure 3A One or more workflows 102 and / or one or more workflows 302 are further described.

[0120] In some embodiments, the workflow corresponding to the first task 320 may include information specifying one or more computing platforms on which the first container 322A and the second container 322B may be executed. In some embodiments, the workflow may not specify a computing platform on which the first container 322A and / or the second container 322B may be executed. In some embodiments, the workflow and / or the information corresponding to the workflow may indicate and / or specify that the first container 322A and the second container 322B may be executed on the same computing platform and simultaneously and / or substantially simultaneously. In some embodiments, the workflow may indicate that the first container 322A and the second container 322B must be executed together (e.g., simultaneously or substantially simultaneously), otherwise they are not executed.

[0121] In some embodiments, the second task 324 may include a third container 326A and a fourth container 326B. In some embodiments, the third container 326A may include a third set of operations, and the fourth container 326B may include a fourth set of operations. In some embodiments, the third container 326A and / or the fourth container 326B may be as described in the present disclosure, for example, with respect to Figure 1 , Figure 2 and Figure 3A Examples of containers are further described and / or illustrated.

[0122] In some embodiments, the second task 324 may correspond to a workflow, such as, for example, Figure 1 and Figure 3A One or more workflows 102 and / or one or more workflows 302 are further described.

[0123] In some embodiments, the workflow corresponding to the second task 324 may include information specifying one or more computing platforms on which the third container 326A and the fourth container 326B may be executed. In some embodiments, the workflow may not specify a computing platform on which the third container 326A and / or the fourth container 326B may be executed. In some embodiments, the workflow and / or the information corresponding to the workflow may indicate and / or specify that the third container 326A and the fourth container 326B may be executed on the same computing platform and simultaneously and / or substantially simultaneously. In some embodiments, the workflow may indicate that the third container 326A and the fourth container 326B must be executed together (e.g., simultaneously or substantially simultaneously), otherwise they are not executed.

[0124] In some embodiments, the scheduler 332 may obtain information corresponding to the first task 320 including the first container 322A and / or the second container 322B and / or the second task 324 including the third container 326A and / or the fourth container 326B. In some embodiments, the scheduler 332 may be, for example, a task described in the present disclosure. Figure 1 and Figure 3A Examples and / or the like of scheduler 104 and / or scheduler 304 are further described and / or illustrated.

[0125] In some embodiments, the scheduler 332 may obtain and / or receive information corresponding to the first task 320 and / or the second task 324. In some embodiments, the first task 320 and the second task 324 may be included in an individual workflow (e.g., the workflow 302). Additionally or alternatively, the first task 320 may be included in a first workflow and the second task 324 may be included in a second workflow, wherein the first workflow and the second workflow may be, for example, a workflow described in the present disclosure. Figure 1 and Figure 3AExamples of workflow 102 and / or workflow 302 are described and / or shown.

[0126] In some embodiments, the scheduler 332 may be configured to schedule one or more tasks (e.g., the first task 320 and / or the second task 324) including one or more corresponding containers (e.g., the first container 322A, the second container 322B, the third container 326A, and / or the fourth container 326B) to be executed using one or more computing platforms. In some embodiments, the scheduler 332 may be configured to schedule tasks and / or containers for execution based on information corresponding to, for example, one or more workflows associated with the first task 320 and / or the second task 324. Additionally or alternatively, the scheduler 328 may specify one or more tasks and / or containers for execution based on the current availability of computing resources.

[0127] In some embodiments, the scheduler 332 may be configured to specify one or more nodes on which tasks and / or corresponding containers may be executed, and the node corresponds to one or more computing platforms. In some embodiments, one or more nodes may include corresponding representations of computing resources and / or processing resources on which one or more tasks, containers and / or operations may be executed. For example, a computing platform may include a GPU, which may include one or more nodes, and each of the one or more nodes may represent resources associated with a GPU that may execute one or more containers. In some embodiments, the scheduler 332 may be configured to specify one or more containers (e.g., a first container 322A, a second container 322B, a third container 326A, and / or a fourth container 326B) to be executed using one or more nodes associated with one or more computing platforms. In some embodiments, the scheduler 332 may be configured to generate one or more schedules, for example, a first plan 334 and / or a second plan 336.

[0128] In some embodiments, the first plan 334 may include the first node 328A and / or the second node 328B. In some embodiments, the scheduler 332 may specify the first container 322A to execute on the first node 328A and specify the third container 326A to execute on the second node 328B. In the case where the first container 322A and the second container 322B and / or the third container 326A and the fourth container 326B must be scheduled together, scheduling the first container 322A and the third container 326A may result in neither the first container 322A nor the third container 326A being executed.

[0129] In contrast, the second plan 336 may include a third node 330A and a fourth node 330B. In some embodiments, the scheduler 332 may specify a third container 326A to execute on the third node 330A, and specify a fourth container 326B to execute on the fourth node 330B. In the case where the third container 326A and the fourth container 326B must be scheduled together, the third container 326A and the fourth container 326B may be executed. In some embodiments, two or more containers that may be scheduled together may be specified in one or more container groups. The container group indicates that the scheduler 332 may wait for computing resources to become available before scheduling the container corresponding to the container group.

[0130] Can Figure 3B Modifications, additions, or omissions may be made without departing from the scope of the present disclosure. For example, the number of tasks, the number of containers corresponding to the respective tasks, the number of schedulers, the number of nodes and plans configured to schedule the execution of one or more tasks and / or one or more containers may vary. The details given and discussed are helpful in explaining and understanding the concepts of the present disclosure and are not meant to be limiting.

[0131] Back to Figure 3A In some embodiments, the co-scheduling module 308 may be configured to schedule multiple tasks to be executed simultaneously or substantially simultaneously. For example, one or more task groups may be scheduled, as described in the present disclosure, for example, regarding Figure 1 and Figure 2 As further described and / or illustrated. For example, the first task and the second task can be designated to be executed simultaneously or substantially simultaneously based on, for example, information corresponding to one or more workflows 302. The first task can be designated to be executed on a first computing platform and the second task can be designated to be executed on a second computing platform. Further continuing with this example, the co-scheduling module 308 can schedule the first task to the first computing platform simultaneously or substantially simultaneously with scheduling the second task to the second computing platform. Additionally or alternatively, the first task and the second task can be designated to be executed on the same computing platform.

[0132] In some embodiments, the scheduler 304 may additionally include a hosting module 310. In some embodiments, the hosting module 310 may be configured to obtain the workflow 302 and / or information corresponding to the workflow 302. Additionally or alternatively, the hosting module 310 may be configured to schedule one or more tasks and / or one or more containers corresponding to the one or more tasks to be executed using a computing platform that may meet network requirements corresponding to the tasks, containers, and / or operations that may be executed.

[0133] In some embodiments, the hosting module 310 can be configured to assign a location key to one or more tasks, containers and / or operations, and the location key can indicate one or more network and / or location attributes. The network and / or location attributes indicate similar physical locations and / or network topologies, so that the tasks, containers and / or operations corresponding to the same location key can have a suitable network topology for execution. For example, the first task and the second task corresponding to a specific workflow 302 can be interdependent, and therefore, the first task and the second task can be executed together. In some embodiments, the first task and the second task can communicate with each other to perform one or more operations. In some embodiments, due to this interdependence, the hosting module 310 can be configured to schedule the first task and the second task to one or more computing platforms, which can include the same or similar network locations, connectivity, etc.

[0134] In some embodiments, the scheduler 304 may additionally include a priority upgrade module 312. In some embodiments, the priority upgrade module 312 may indicate one or more priority levels that may be attached to one or more tasks, containers, and / or operations to be executed. In some embodiments, the priority level may indicate which tasks, containers, and / or operations may be executed first. For example, the second workflow 302 may be received after the first workflow 302. Continuing with this example, the tasks corresponding to the second workflow 302 may include a high priority so that the tasks corresponding to the second workflow 302 may be executed before one or more other tasks corresponding to the first workflow 302.

[0135] In some embodiments, priority upgrade module 312 can be configured to upgrade one or more priority levels corresponding to task, container and / or operation, and these priority levels can wait for the time amount to be scheduled based on task, container and / or operation.For example, one or more tasks, container and / or operation corresponding to first workflow 302 can be included in queue.In some embodiments, scheduler 304 can be configured to schedule tasks, container and / or operation in queue based on the current availability of information and / or computing resources corresponding to workflow 302.Continue this example, when the time of specific task in queue exceeds predetermined time amount, priority upgrade module 312 can upgrade one or more priority levels corresponding to specific task, so that specific task can be scheduled before one or more other tasks, container and / or operation.In some cases, upgrading specific task in this way can cause waiting for enough computing resources to become available for executing this specific task.

[0136] In some embodiments, upgrading one or more tasks, containers, and / or operations can improve the efficient utilization of computing resources and reduce the probability that the aforementioned specific tasks, containers, operations, and / or groups can remain in the queue, thereby allowing tasks, containers, and / or groups corresponding to the same workflow 302 to be scheduled.

[0137] In some embodiments, scheduler 304, co-scheduling module 308, hosting module 310, and / or priority escalation module 312 may be included in runtime system 314. In some embodiments, determinations made using scheduler 304, co-scheduling module 308, hosting module 310, and / or priority escalation module 312 may be made in real time or substantially real time based on information corresponding to workflow 302 and current availability of computing resources corresponding to one or more computing platforms.

[0138] In some embodiments, the scheduler 304, the co-scheduling module 308, the hosting module 310, and / or the priority upgrade module 312 can facilitate the generation of one or more execution plans 306. In these and other embodiments, the one or more execution plans 306 can be the same and / or similar to one or more execution plans 106 and the example execution plan 200, which can be described in the present disclosure, for example, with respect to Figure 1 and Figure 2 Further description and / or illustration.

[0139] Can Figure 3A Modifications, additions, or omissions may be made without departing from the scope of the present disclosure. For example, the number of workflows 302, the number and / or type of systems corresponding to schedulers 304, the number of modules (e.g., co-scheduling modules 308), the number of managed modules 310, the number of priority escalation modules 312. Additionally or alternatively, the number of execution plans 306 may vary. The details presented and discussed are helpful in explaining and understanding the concepts of the present disclosure and are not meant to be limiting.

[0140] Figure 4 An example environment 400 is shown for generating one or more outputs 410 based on executing one or more workflows corresponding to one or more execution plans 406 according to one or more embodiments of the present disclosure. In these and other embodiments, one or more execution plans 406 may be associated with, for example, Figure 1 and Figure 3AThe one or more execution plans 106 and / or one or more execution plans 306 described and / or illustrated further are the same and / or similar. In some embodiments, the one or more execution plans 406 can be generated and / or created in real time or substantially real time based on the workflow, information corresponding to the workflow (e.g., execution information), and / or availability of resources associated with one or more computing platforms. In some embodiments, the one or more execution plans 406 can be sent, transmitted, and / or otherwise communicated to the execution module 408.

[0141] In some embodiments, the execution module 408 can be included in the runtime system 420 so that one or more determinations and / or operations performed using the execution module 408 can be performed in real time or substantially in real time. In these and other embodiments, the runtime system 420 corresponding to the environment 400 can be similar to the runtime system 420 described in the present disclosure, for example, with respect to Figure 3A The runtime system 314 described and / or illustrated further may be the same and / or similar.

[0142] In some embodiments, the execution module 408 can be configured to receive and / or otherwise obtain one or more execution plans 406. Additionally or alternatively, the execution module 408 can be configured to generate one or more outputs 410 based on the execution of one or more tasks, containers, and / or operations corresponding to the one or more execution plans 406. In some embodiments, the execution module 408 can include one or more modules that can be configured to generate one or more outputs 410. For example, Figure 4 As shown, the execution module 408 may include an initialization module 412 , a timestamp module 414 , an error module 416 , and / or a termination module 418 .

[0143] In some embodiments, the execution module, initialization module 412, timestamp module 414, error module 416 and / or termination module 418 may include code and routines configured to allow the computing system to perform one or more operations. Additionally or alternatively, these modules may be implemented using hardware, including one or more processors, CPUs, graphics processing units (GPUs), data processing units (DPUs), parallel processing units (PPUs), microprocessors (e.g., for executing or controlling the execution of one or more operations), field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), accelerators (e.g., deep learning accelerators (DLAs)), and / or other processor types. In these and other embodiments, the corresponding modules may be implemented using a combination of hardware and software. In the present disclosure, operations described as being performed by any of the corresponding modules may include operations that the modules may instruct the corresponding computing system to perform. In these or other embodiments, the execution module, initialization module 412, timestamp module 414, error module 416 and / or termination module 418 may be implemented by one or more computing devices, such as with respect to Figures 6A-6D , Figure 7 and / or Figure 8 those described in further detail.

[0144] The initialization module 412 may be configured to initialize execution of one or more tasks, containers, operations, etc. according to the one or more execution plans 406. In some embodiments, the initialization module 412 may instruct one or more computing platforms (e.g., one or more nodes associated with the one or more computing platforms) to execute one or more tasks, containers, and / or operations corresponding to the one or more execution plans 406.

[0145] In some embodiments, in response to the initialization module 412 instructing one or more computing platforms to execute one or more tasks, containers and / or operations, the execution module 408 can be configured to monitor the execution of the tasks, containers and / or operations using, for example, the timestamp module 414, the error module 416 and / or the termination module 418.

[0146] In some embodiments, timestamp module 414 can be configured to assign timestamps to one or more tasks, containers, and / or operations being executed. In some embodiments, one or more timeouts corresponding to one or more workflows can be determined by assigning corresponding timestamps to the corresponding tasks that can be executed. For example, an individual task can correspond to a specific workflow. Continuing with this example, the specific workflow can indicate the time at which the individual task can be terminated thereafter. In some embodiments, in order to determine the amount of time that can have passed after initializing execution, timestamp module 414 can assign timestamps to tasks when execution has been initialized.

[0147] In some embodiments, the timestamp module 414 may be configured to generate one or more queries at runtime to determine whether a task, container, operation, etc. has completed execution. In some embodiments, the timestamp module 414 may generate queries at a specified timeout period corresponding to the corresponding task, container, operation, etc. In some embodiments, the timestamp module 414 may determine that the task, container, operation, etc. may still be running and / or executing, or that the task, container, operation, etc. may have completed execution. In the event that the task, container, operation, etc. has not completed execution after the specified timeout period, the timestamp module 414 may be configured to communicate this information to the error module 416 and / or the termination module 418.

[0148] In some embodiments, the error module 416 may be configured to determine whether one or more errors have occurred during execution. In some embodiments, the error module 416 may receive data and / or information from the timestamp module 414 that may indicate that a task, container, operation, etc. may still be running and / or executing after a corresponding timeout period. Additionally or alternatively, the error module 416 may be configured to generate one or more queries that may elicit information associated with the execution of one or more tasks from the corresponding computing platform.

[0149] In some embodiments, the query may elicit error information associated with executing one or more tasks, containers, operations, and the like. For example, errors may include errors in code (e.g., syntax errors, logic errors, and the like), network errors, file input / output operation errors, security errors, memory errors, and / or other errors that may be encountered during the execution of one or more tasks, containers, and / or operations. Additionally or alternatively, the error module 416 may generate error information as part of the output 410. For example, the error module 416 may generate a log corresponding to the task, container, operation, and the like, wherein the log includes information corresponding to the one or more errors, such as what the one or more errors may be, potential causes of the one or more errors, possible consequences of the one or more errors to one or more other systems, subsystems, machines, and the like. In some embodiments, the error module 416 may be configured to transmit the error information to the termination module 418.

[0150] Termination module 418 can be configured to cancel and / or suspend the execution of one or more tasks, container and / or operation based on the error information detected by using, for example, error module 416. In certain embodiments, termination module 418 can stop and / or suspend the execution of corresponding tasks, container, operation etc., no matter how the error type. Additionally or alternatively, termination module 418 can stop and / or suspend the execution of corresponding tasks, container and / or operation based on what error is detected. For example, the task that performs beyond the specified timeout period can be stopped, and the task that comprises one or more slight grammatical errors can continue to be executed. In certain embodiments, termination module 418 can transmit the information associated with the execution of stopping and / or suspending one or more tasks, container, operation etc. as a part of output 410.

[0151] In some embodiments, without detecting one or more errors, execution module 408, initialization module 412, timestamp module 414, error module 416, and / or termination module 418 may indicate that a task, container, operation, etc. may have completed execution without errors. Additionally, in some embodiments, results corresponding to tasks, containers, operations, etc. that may have been executed may be included in output 410. In these and other embodiments, output 410 may be similar to those described in the present disclosure, for example, with respect to Figure 1 The output 110 described and / or illustrated further is the same and / or similar.

[0152] Can Figure 4 Modifications, additions, or omissions may be made without departing from the scope of the present disclosure. For example, the number of execution plans 406 may vary. Additionally or alternatively, workflow execution module 408 may correspond to one or more different types of systems, subsystems, machines, etc. In some embodiments, one or more other modules, systems, subsystems, etc. may be used to generate information about the workflow execution plan. Figure 4 The functionality described, particularly with respect to the initialization module 412, the timestamp module 414, the error module 416, and / or the termination module 418. The details presented and discussed are intended to aid in explaining and understanding the concepts of the present disclosure and are not meant to be limiting.

[0153] Figure 5 is a flow chart illustrating a method 500 for initiating execution of one or more tasks corresponding to a corresponding workflow according to one or more embodiments of the present disclosure. The method 500 may include one or more blocks 502, 504, and 506. Although illustrated as discrete blocks, operations associated with one or more blocks of the method 500 may be divided into additional blocks, combined into fewer blocks, or eliminated, depending on the specific implementation.

[0154] In some embodiments, method 500 may include box 502. At box 502, execution information corresponding to the first task and the second task may be obtained. In some embodiments, the first task obtained may include one or more first containers, the first container may include a first operation, and the second task may include one or more second containers, the second container may include a second operation. In some embodiments, the execution information may identify one or more first dependencies that may correspond to the first task. In some embodiments, one or more second dependencies may correspond to the second task. In some embodiments, the execution information may be used to identify a first computing platform on which the first task is to be executed, and additionally identify a second computing platform on which the second task may be executed. In some embodiments, the first computing platform may be different from the second computing platform. In some embodiments, the first computing platform and / or the second computing platform may include one or more of a CPU, a GPU, a TPU, an FPGA, a DPU, a PPU, an ASIC, and / or a DLA.

[0155] At block 504, a time to initialize execution of the first task and the second task may be determined in real time or substantially in real time. In some embodiments, the determination of the time to initialize execution may be determined based on execution information corresponding to the first task and the second task. In some embodiments, the execution information may identify one or more first dependencies corresponding to the first task and one or more second dependencies corresponding to the second task. In some embodiments, the execution information may additionally include a first computing platform on which the first task is to be executed and a second computing platform on which the second task is to be executed, wherein the first computing platform may be different from the second computing platform.

[0156] In some embodiments, the execution information may additionally include current availability of computing resources that may correspond to the first computing platform and the second computing platform. In some embodiments, the current availability of computing resources may allow one or more tasks, containers, and / or operations to be scheduled for execution in real time or substantially real time.

[0157] In some embodiments, the execution information may additionally include a first timeout period that may correspond to the first task and a second timeout period that may correspond to the second task. In some embodiments, the first timeout period and the second timeout period may respectively indicate a time period after which the execution of the first operation and the second operation is terminated. In some embodiments, the first timeout period and the second timeout period may be the same.

[0158] At block 506, execution of the first task on the first computing platform and the second task on the second computing platform may be initialized. In some embodiments, execution of the one or more first containers and the one or more second containers may be interdependent.

[0159] Method 500 and / or one or more operations included in method 500 may be modified, added, or omitted without departing from the scope of the present disclosure. For example, operations corresponding to method 500 may be implemented in a different order. Additionally or alternatively, two or more operations may be performed simultaneously. Furthermore, the operations and actions outlined are provided only as examples, and some operations and actions may be optional, combined into fewer operations and actions, or expanded into additional operations and actions without departing from the essence of the described embodiments.

[0160] Example Autonomous Vehicle

[0161] Fig. 6A 6 is a diagram of an exemplary autonomous vehicle 600 according to some embodiments of the present disclosure. Autonomous vehicle 600 (or referred to herein as "vehicle 600") may include, but is not limited to, passenger vehicles, such as cars, trucks, buses, first response vehicles, shuttles, electric or motorized bicycles, motorcycles, fire trucks, police cars, ambulances, boats, engineering vehicles, submarines, drones, and / or other types of vehicles (e.g., vehicles that are unmanned and / or accommodate one or more passengers). Autonomous vehicles are generally described according to the level of automation, as defined by the National Highway Traffic Safety Administration (NHTSA) of the U.S. Department of Transportation, and the Society of Automotive Engineers (SAE) "Classification and Definition of Terms Related to Driving Automation Systems for Road Motor Vehicles" (Standard No.: J3016-201806, published on June 15, 2018, Standard No.: J3016-201609, published on September 30, 2016, and previous and future versions of this standard). Vehicle 600 may implement functions according to one or more of the levels 3 to 5 of autonomous driving levels. The vehicle 600 may be capable of having functionality consistent with one or more of Levels 1 to 5 of autonomous driving levels. For example, depending on the embodiment, the vehicle 600 may have driver assistance (Level 1), partial automation (Level 2), conditional automation (Level 3), high automation (Level 4), and / or full automation (Level 5). The term "autonomy" as used herein may include any and / or all types of autonomy of the vehicle 600 or other machine, such as fully autonomous, highly autonomous, conditionally autonomous, partially autonomous, providing assisted autonomy, semi-autonomous, primarily autonomous, or other designations.

[0162] The vehicle 600 may include a chassis, a body, wheels (e.g., 2, 4, 6, 8, 18, etc.), tires, axles, and other components of the vehicle. The vehicle 600 may include a propulsion system 650, such as an internal combustion engine, a hybrid power plant, an all-electric engine, and / or other propulsion system types. The propulsion system 650 may be connected to a drive train of the vehicle 600, which may include a transmission, to achieve propulsion of the vehicle 600. The propulsion system 650 may be controlled in response to receiving a signal from a throttle / accelerator 652.

[0163] When propulsion system 650 is in operation (e.g., when the vehicle is in motion), a steering system 654 including a steering wheel may be used to guide vehicle 600 (e.g., along a desired path or route). Steering system 654 may receive signals from steering actuator 656. For fully automated (Level 5) functionality, a steering wheel may be optional.

[0164] Brake sensor system 646 may be used to operate vehicle brakes in response to signals received from brake actuator 648 and / or brake sensors.

[0165] The one or more controllers 636 may include one or more CPUs, system on chip (SoC) 604 ( Figure 6C ) and / or one or more GPUs, may provide signals (e.g., representing commands) to one or more components and / or systems of the vehicle 600. For example, the one or more controllers may send signals to operate vehicle brakes via one or more brake actuators 648, operate the steering system 654 via one or more steering actuators 656, and / or operate the propulsion system 650 via one or more throttles / accelerators 652. The one or more controllers 636 may include one or more on-board (e.g., integrated) computing devices (e.g., supercomputers) that process sensor signals and output operational commands (e.g., signals representing commands) to enable autonomous driving and / or assist a human driver in driving the vehicle 600. The one or more controllers 636 may include a first controller 636 for autonomous driving functions, a second controller 636 for functional safety functions, a third controller 636 for artificial intelligence functions (e.g., computer vision), a fourth controller 636 for infotainment functions, a fifth controller 636 for redundancy in emergency situations, and / or other controllers. In some examples, a single controller 636 may handle two or more of the above functions, two or more controllers 636 may handle a single function, and / or any combination thereof.

[0166] The one or more controllers 636 may provide signals for controlling one or more components and / or systems of the vehicle 600 in response to sensor data (eg, sensor inputs) received from one or more sensors. Sensor data may be received from, for example, but not limited to, one or more global navigation satellite system sensors 658 (e.g., one or more global positioning system sensors), one or more RADAR sensors 660, one or more ultrasonic sensors 662, one or more LIDAR sensors 664, one or more inertial measurement unit (IMU) sensors 666 (e.g., one or more accelerometers, one or more gyroscopes, one or more magnetic compasses, magnetometers, etc.), one or more microphones 696, one or more stereo cameras 668, one or more wide-angle cameras 670 (e.g., fisheye cameras), one or more infrared cameras 672, one or more surround cameras 674 (e.g., 360 degree cameras), one or more long-range and / or medium frequency cameras 698, one or more speed sensors 644 (e.g., for measuring the speed of vehicle 600), one or more vibration sensors 642, one or more steering sensors 640, one or more brake sensors 646 (e.g., as part of brake sensor system 646), and / or other sensor types.

[0167] One or more of the controllers 636 may receive input (e.g., represented by input data) from the instrument cluster 632 of the vehicle 600 and provide output (e.g., represented by output data, display data, etc.) via a human machine interface (HMI) display 634, an audio annunciator, a speaker, etc. and / or via other components of the vehicle 600. The output may include information such as vehicle speed, velocity, time, map data (e.g., Figure 6C The HMI display 634 may include information such as the HD map 622 of the vehicle 600, location data (e.g., the location of the vehicle 600, such as the location on the map), directions, the locations of other vehicles (e.g., occupancy grid), and information about objects and object states sensed by one or more controllers 636. For example, the HMI display 634 may display information about the presence of one or more objects (e.g., street signs, warning signs, traffic light changes, etc.), and / or information about driving operations that the vehicle has performed, is performing, or will perform (e.g., change lanes now, take exit 34B in two miles, etc.).

[0168] The vehicle 600 also includes a network interface 624 that can communicate via one or more networks using one or more wireless antennas 626 and / or a modem. For example, the network interface 624 can be capable of communicating via LTE, WCDMA, UMTS, GSM, CDMA2000, etc. The one or more wireless antennas 626 can also enable communication between objects in the environment (e.g., vehicles, mobile devices, etc.) using one or more local area networks (e.g., Bluetooth, Bluetooth LE, Z-wave, ZigBee, etc. and / or low power wide area networks (LPWANs) such as LoRaWAN, SigFox, etc.).

[0169] Figure 6B According to some embodiments of the present disclosure Fig. 6A 600. The cameras and respective fields of view are an example embodiment and are not intended to be limiting. For example, additional and / or alternative cameras may be included and / or the cameras may be located at different locations on the vehicle 600.

[0170] The camera type of the camera may include, but is not limited to, a digital camera, which may be applicable to components and / or systems of the vehicle 600. One or more cameras may operate at automotive safety integrity level (ASIL) B and / or other ASILs. According to an embodiment, the camera type may have any image capture rate, such as 60 frames per second (fps), 120 frames, 240 frames, etc. The camera may use a rolling shutter, a global shutter, another type of shutter, or a combination thereof. In some examples, the color filter array may include a red transparent (RCCC) color filter array, a red transparent blue (RCCB) color filter array, a red, blue, green transparent (RBGC) color filter array, a Foveon X3 color filter array, a Bayer sensor (RGGB) color filter array, a monochrome sensor color filter array, and / or other types of color filter arrays. In some embodiments, a clear pixel camera, such as a camera with an RCCC, RCCB, and / or RBGC color filter array, may be used in an effort to improve light sensitivity.

[0171] In some examples, one or more cameras can be used to perform advanced driver assistance system (ADAS) functions (e.g., as part of a redundant or fail-safe design). For example, a multifunctional single camera can be installed to provide functions such as lane departure warning, traffic sign assistance, and intelligent headlight control. One or more cameras (e.g., all cameras) can record and provide image data (e.g., video) simultaneously.

[0172] One or more cameras may be mounted in a mounting assembly, such as a custom designed (3D printed) assembly, to cut out stray light and in-vehicle reflections (e.g., dashboard reflections off of windshield mirrors) that may interfere with the camera's ability to capture image data. With respect to the wing mirror mounting assembly, the wing mirror assembly may be custom 3D printed so that the camera mounting plate matches the shape of the wing mirror. In some examples, one or more cameras may be integrated into the wing mirror. For side view cameras, one or more cameras may also be integrated into the four pillars at each corner of the cabin.

[0173] A camera (e.g., a front-facing camera) having a field of view that includes a portion of the environment in front of the vehicle 600 can be used for a surround view to help identify the forward path and obstacles, as well as assist in providing information critical to generating an occupancy grid and / or determining a preferred vehicle path with the help of one or more controllers 636 and / or a controlling SOC. A front-facing camera can be used to perform many of the same ADAS functions as LIDAR, including emergency braking, pedestrian detection, and collision avoidance. A forward-facing camera can also be used for ADAS functions and systems, including lane departure warning ("LDW"), automatic cruise control ("ACC"), and / or other functions, such as traffic sign recognition.

[0174] A variety of cameras may be used in the front-facing configuration, including, for example, a monocular camera platform including a CMOS (complementary metal oxide semiconductor) color imager. Another example may be a wide-angle camera 670 that may be used to sense objects (e.g., pedestrians, crossing traffic, or bicycles) that enter the field of view from the periphery. Although Figure 6B Only one wide-angle camera is shown, but there may be any number of wide-angle cameras 670 on the vehicle 600. In addition, remote cameras 698 (e.g., a long-angle stereo camera pair) can be used for depth-based object detection, especially for objects for which a neural network has not yet been trained. One or more remote cameras 698 can also be used for object detection and classification and basic target tracking.

[0175] One or more stereo cameras 668 may also be included in the front configuration. One or more stereo cameras 668 may include an integrated control unit including an expandable processing unit that may provide programmable logic ("FPGA") and a multi-core microprocessor with an integrated CAN or Ethernet interface on a single chip. The unit may be used to generate a 3D map of the vehicle environment, including distance estimates for all points in the image. One or more alternative stereo cameras 668 may include a compact stereo vision sensor that may include two camera lenses (one on each side) and an image processing chip that may measure the distance from the vehicle to the target object and use the generated information (e.g., metadata) to activate automatic emergency braking and lane departure warning functions. In addition to the stereo cameras described herein, or alternatively, other types of stereo cameras 668 may be used.

[0176] Cameras with a field of view that includes portions of the surroundings to the sides of the vehicle 600 (e.g., side view cameras) can be used for surround views, providing information for creating and updating occupancy grids and generating side collision warnings. For example, one or more surround cameras 674 (e.g., Figure 6B Four surround cameras 674 (shown) may be positioned on the vehicle 600. The one or more surround cameras 674 may include one or more wide angle cameras 670, one or more fisheye cameras, one or more 360 ​​degree cameras, etc. For example, four fisheye cameras may be located at the front, rear, and sides of the vehicle. In an alternative arrangement, the vehicle may use three surround cameras 674 (e.g., left, right, and rear), and may utilize one or more other cameras (e.g., a front-facing camera) as a fourth surround view camera.

[0177] A camera having a field of view that includes a portion of the environment behind the vehicle 600 (e.g., a rear view camera) can be used for parking assistance, surround view, rear-end collision warning, and creating and updating an occupancy grid. A variety of cameras can be used, including but not limited to cameras that are also suitable as front-facing cameras (e.g., one or more long-range and / or mid-range cameras 698, one or more stereo cameras 668, one or more infrared cameras 672, etc.), as described herein.

[0178] Figure 6C According to some embodiments of the present disclosure Fig. 6ABlock diagram of an exemplary system architecture of an exemplary autonomous vehicle 600. It should be understood that this arrangement and other arrangements described herein are set forth only as examples. In addition to the arrangements and elements shown, other arrangements and elements (e.g., machines, interfaces, functions, sequences, functional groupings, etc.) may be used, or other arrangements and elements may be used in place of the arrangements and elements shown, and some elements may be omitted entirely. In addition, many of the elements described herein are functional entities that may be implemented as discrete or distributed components, or in combination with other components, and may be implemented in any suitable combination and location. The various functions described herein as being performed by an entity may be performed by hardware, firmware, and / or software. For example, the various functions may be performed by a processor executing instructions stored in a memory.

[0179] Figure 6C Each component, feature, and system of vehicle 600 is connected via bus 602. Bus 602 may include a controller area network (CAN) data interface (also referred to herein as a "CAN bus"). CAN can be a network within vehicle 600 that helps control various features and functions of vehicle 600, such as the drive of brakes, acceleration, braking, steering, windshield wipers, etc. The CAN bus can be configured to have dozens or even hundreds of nodes, each with its own unique identifier (e.g., CAN ID). The CAN bus can be read to find steering wheel angle, ground speed, engine speed per minute (RPM), button position, and / or other vehicle status indicator lights. The CAN bus can comply with ASIL B standards.

[0180] Although bus 602 is described herein as a CAN bus, this is not intended to be limiting. For example, in addition to or from the CAN bus, FlexRay and / or Ethernet can also be used. In addition, although a single line is used to represent bus 602, this is not intended to be limiting. For example, there can be any number of buses 602, which may include one or more CAN buses, one or more FlexRay buses, one or more Ethernet buses, and / or one or more other types of buses using different protocols. In some examples, two or more buses 602 can be used to perform different functions, and / or can be used for redundancy. For example, a first bus 602 can be used for collision avoidance functions, and a second bus 602 can be used for drive control. In any example, each bus 602 can communicate with any component of the vehicle 600, and two or more buses 602 can communicate with the same component. In some examples, each SoC 604, each controller 636, and / or each computer in the vehicle can access the same input data (e.g., input from sensors of the vehicle 600), and can be connected to a common bus, such as a CAN bus.

[0181] The vehicle 600 may include one or more controllers 636, as described herein with respect to Fig. 6A The controller 636 can be used for various functions. One or more controllers 636 can be coupled to any of the various other components and systems of the vehicle 600, and can be used to control the vehicle 600, the artificial intelligence of the vehicle 600, the infotainment of the vehicle 600, etc.

[0182] The vehicle 600 may include one or more systems on a chip (SoC) 604. The SoC 604 may include one or more CPUs 606, one or more GPUs 608, one or more processors 610, one or more caches 612, one or more accelerators 614, one or more data stores 616, and / or other components and features not shown. The one or more SoCs 604 may be used to control the vehicle 600 in various platforms and systems. For example, the one or more SoCs 604 may be combined with an HD map 622 in a system (e.g., a system of the vehicle 600), and the HD map 622 may be downloaded from one or more servers (e.g., Fig.6D server 678) to obtain map refreshes and / or updates.

[0183] One or more CPUs 606 may include a CPU cluster or CPU complex (or referred to herein as a "CCPLEX"). One or more CPUs 606 may include multiple cores and / or L2 caches. For example, in some embodiments, one or more CPUs 606 may include eight cores in a coherent multiprocessor configuration. In some embodiments, one or more CPUs 606 may include four dual-core clusters, each with a dedicated L2 cache (e.g., a 2MB L2 cache). One or more CPUs 606 (e.g., CCPLEX) may be configured to support simultaneous cluster operations so that any combination of CPU 606 clusters is active at any given time.

[0184] One or more CPUs 606 may implement power management capabilities including one or more of the following features: individual hardware blocks may be automatically clock gated when idle to save dynamic power; each core clock may be gated when the core is not actively executing instructions due to executing WFI / WFE instructions; each core may be independently power gated; each core cluster may be independently clock gated when all cores are clock gated or power gated; and / or each core cluster may be independently power gated when all cores are power gated. One or more CPUs 606 may further implement an enhanced algorithm for managing power states, where allowed power states and expected wakeup times are specified, and hardware / microcode determines the optimal power state to enter for the core, cluster, and CCPLEX. The processing core may support a simplified power state entry sequence in software and offload work to microcode.

[0185] One or more GPUs 608 may include an integrated GPU (or referred to herein as an "iGPU"). GPU 608 may be programmable and may be efficient for parallel workloads. In some examples, one or more GPUs 608 may use an enhanced tensor instruction set. One or more GPUs 608 may include one or more stream microprocessors, each of which may include a level 1 cache (e.g., a level 1 cache with at least 96KB storage capacity), and two or more stream microprocessors may share a level 2 cache (e.g., a level 2 cache with 512KB storage capacity). In some embodiments, one or more GPUs 608 may include at least eight stream microprocessors. One or more GPUs 608 may use one or more computing application programming interfaces (APIs). In addition, one or more GPUs 608 may use one or more parallel computing platforms and / or programming models (e.g., NVIDIA's CUDA).

[0186] One or more GPUs 608 may be power optimized to achieve optimal performance in automotive and embedded use cases. For example, one or more GPUs 608 may be manufactured on fin field effect transistors (FinFETs). However, this is not intended to be limiting, and one or more GPUs 608 may be manufactured using other semiconductor manufacturing processes. Each streaming microprocessor may incorporate multiple mixed precision processing cores that are divided into multiple blocks. For example, but not limited to, 64 PF32 cores and 32 PF64 cores may be divided into four processing blocks. In such an example, 16 FP32 cores, 8 FP64 cores, 16 INT32 cores, two mixed precision NVIDIA tensor cores for deep learning matrix algorithms, L0 instruction caches, warp schedulers, dispatch units, and / or 64KB register files may be allocated to each processing block. In addition, the streaming microprocessor may include independent parallel integer and floating point data paths to provide efficient execution of workloads through mixed calculations and addressing calculations. The streaming microprocessor may include independent thread scheduling capabilities to achieve finer-grained synchronization and collaboration between parallel threads. A streaming microprocessor may include a combined L1 data cache and shared memory unit to increase performance while simplifying programming.

[0187] One or more GPUs 608 may include high bandwidth memory (HBM) and / or a 16GB HBM2 memory subsystem to provide about 900GB / sec peak memory bandwidth in some examples. In some examples, synchronous graphics random access memory (SGRAM), such as fifth generation graphics double data rate synchronous random access memory (GDDR5), is used in addition to or alternatively from the HBM memory.

[0188] The fifth generation GPU 608 may include unified memory technology, which includes access counters to allow more accurate migration of memory pages to the processor that most frequently accesses them, thereby improving the efficiency of sharing memory ranges between processors. In some examples, address translation service (ATS) support may be used to allow one or more GPUs 608 to directly access one or more CPU 606 page tables. In such an example, when one or more GPU 608 memory management units (MMUs) experience a miss, an address translation request may be sent to one or more CPUs 606. In response, one or more CPUs 606 may look up the virtual to physical mapping of the address in its page table and send the translation back to one or more GPUs 608. Therefore, unified memory technology may allow a single unified virtual address space for the memory of both one or more CPUs 606 and one or more GPUs 608, thereby simplifying one or more GPU 608 programming and porting applications to one or more GPUs 608.

[0189] In addition, one or more GPUs 608 may include access counters that can track how often one or more GPUs 608 access the memory of other processors. The access counters can help ensure that memory pages are moved to the physical memory of the processor that accesses the page most frequently.

[0190] The one or more SoCs 604 may include any number of caches 612, including the caches 612 described herein. For example, the one or more caches 612 may include an L3 cache available to both the one or more CPUs 606 and the one or more GPUs 608 (e.g., connected to both the one or more CPUs 606 and the one or more GPUs 608). The cache 612 may include a write-back cache that can track the state of a line, such as by using a cache coherence protocol (e.g., MEI, MESI, MSI, etc.). Although smaller cache sizes may be used, according to an embodiment, the L3 cache may include 4MB or more.

[0191] One or more SoCs 604 may include one or more arithmetic logic units (ALUs) that may be used to perform processing related to various tasks or operations of the vehicle 600, such as processing a DNN. In addition, one or more SoCs 604 may include one or more floating point units (FPUs) or other math coprocessor or digital coprocessor types for performing mathematical operations within the system. For example, one or more SoCs 604 may include one or more FPUs integrated as execution units within the CPU 606 and / or GPU 608.

[0192] One or more SoCs 604 may include one or more accelerators 614 (e.g., hardware accelerators, software accelerators, or a combination thereof). For example, one or more SoCs 604 may include a hardware acceleration cluster that may include optimized hardware accelerators and / or large on-chip memory. Large on-chip memory (e.g., 4MB of SRAM) can enable the hardware acceleration cluster to accelerate neural networks and other calculations. The hardware acceleration cluster can be used to supplement one or more GPUs 608 and unload some tasks of one or more GPUs 608 (e.g., freeing up more cycles of one or more GPUs 608 to perform other tasks). For example, one or more accelerators 614 can be used for target workloads that are stable enough to be suitable for acceleration (e.g., perception, convolutional neural networks (CNNs), etc.). The term "CNN" used herein may include all types of CNNs, including regional or regional convolutional neural networks (RCNNs) and fast RCNNs (e.g., for target detection).

[0193] One or more accelerators 614 (e.g., a hardware acceleration cluster) may include one or more deep learning accelerators (DLAs). One or more DLAs may include one or more tensor processing units (TPUs) that may be configured to provide an additional 10 trillion operations per second for deep learning applications and reasoning. A TPU may be an accelerator configured to perform image processing functions (e.g., for CNN, RCNN, etc.) and optimized for them. One or more DLAs may also be optimized for specific neural network types and floating point operations and reasoning. The design of one or more DLAs may provide higher per-millimeter performance than a general-purpose GPU, and greatly exceed the performance of a CPU. One or more TPUs may perform multiple functions, including single-instance convolution functions, for example, supporting INT8, INT16, and FP16 data types for features and weights, and post-processor functions.

[0194] One or more DLAs may quickly and efficiently execute neural networks, particularly CNNs, on processed or unprocessed data for any of a variety of functions, including but not limited to: a CNN for object recognition and detection using data from a camera sensor; a CNN for distance estimation using data from a camera sensor; a CNN for emergency vehicle detection and identification and detection using microphone data; a CNN for facial recognition and vehicle owner identification using data from a camera sensor; and / or a CNN for safety and / or security-related events.

[0195] One or more DLAs can perform any function of one or more GPUs 608, for example, by using an inference accelerator, the designer can perform any function for one or more DLAs or one or more GPUs 608. For example, the designer can focus the processing of CNNs and floating point operations on one or more DLAs and leave other functions to one or more GPUs 608 and / or one or more other accelerators 614.

[0196] One or more accelerators 614 (e.g., a hardware acceleration cluster) may include a programmable vision accelerator (PVA), which may also be referred to herein as a computer vision accelerator. One or more PVAs may be designed and configured to accelerate computer vision algorithms for advanced driver assistance systems (ADAS), autonomous driving, and / or augmented reality (AR) and / or virtual reality (VR) applications. One or more PVAs may provide a balance between performance and flexibility. For example, each PVA may include, for example, but not limited to, any number of reduced instruction set computer (RISC) cores, direct memory access (DMA), and / or any number of vector processors.

[0197] The RISC core may interact with an image sensor (e.g., an image sensor of any camera described herein), one or more image signal processors, and the like. Each RISC core may include any number of memories. The RISC core may use any of a variety of protocols, depending on the embodiment. In some examples, the RISC core may execute a real-time operating system (RTOS). The RISC core may be implemented using one or more integrated circuit devices, application specific integrated circuits (ASICs), and / or memory devices. For example, the RISC core may include an instruction cache and / or a tightly coupled RAM.

[0198] The DMA may enable the PVA component to access system memory independently of the one or more CPUs 606. The DMA may support any functionality for optimizing the PVA, including, but not limited to, supporting multi-dimensional addressing and / or circular addressing. In some examples, the DMA may support addressing in up to six or more dimensions, which may include block width, block height, block depth, horizontal block stepping, vertical block stepping, and / or depth stepping.

[0199] A vector processor may be a programmable processor designed to efficiently and flexibly execute programming of computer vision algorithms and provide signal processing capabilities. In some examples, the PVA may include a PVA core and two vector processing subsystem partitions. The PVA core may include a processor subsystem, one or more DMA engines (e.g., two DMA engines), and / or other peripherals. The vector processing subsystem may operate as the main processing engine of the PVA and may include a vector processing unit (VPU), an instruction cache, and / or a vector memory (e.g., VMEM). The VPU core may include a digital signal processor, such as a single instruction, multiple data (SIMD), a very long instruction word (VLIW) digital signal processor. The combination of SIMD and VLIW can improve throughput and speed.

[0200] Each vector processor may include an instruction cache and may be coupled to a dedicated memory. As a result, in some examples, each vector processor may be configured to execute independently of the other vector processors. In other examples, the vector processors included in a particular PVA may be configured to employ data parallelism. For example, in some embodiments, multiple vector processors included in a single PVA may execute the same computer vision algorithm on different regions of an image. In other examples, the vector processors included in a particular PVA may execute different computer vision algorithms simultaneously on the same image, or even execute different algorithms on sequence images or portions of images. Among other things, any number of PVAs may be included in a hardware acceleration cluster, and any number of vector processors may be included in each PVA. In addition, the PVA may include additional error correction code (ECC) memory to enhance overall system security.

[0201] One or more accelerators 614 (e.g., a hardware acceleration cluster) may include a computer vision network on-chip and SRAM to provide high bandwidth, low latency SRAM for one or more accelerators 614. In some examples, the on-chip memory may include at least 4MB of SRAM, including, but not limited to, eight field-configurable memory blocks, accessible by the PVA and DLA. Each pair of memory blocks may include an advanced peripheral bus (APB) interface, configuration circuits, controllers, and multiplexers. Any type of memory may be used. The PVA and DLA may access the memory via a trunk that provides high-speed memory access to the PVA and DLA. The trunk may include an on-chip computer vision network that interconnects the PVA and DLA to the memory (e.g., using APB).

[0202] The on-chip computer vision network may include an interface that determines that both the PVA and DLA provide ready and valid signals before transmitting any control signals / addresses / data. Such an interface may provide independent phases and independent channels for transmitting control signals / addresses / data, as well as burst communication for continuous data transmission. Such an interface may comply with ISO26262 or IEC 61508 standards, although other standards and protocols may also be used.

[0203] In some examples, one or more SoCs 604 may include a real-time ray tracing hardware accelerator, as described in U.S. Patent Application No. 16 / 101,232 filed on August 10, 2018. The real-time ray tracing hardware accelerator may be used to quickly and efficiently determine the position and range of objects (e.g., within a world model), generate real-time visual simulations for radar signal interpretation, sound propagation synthesis and / or analysis, sonar system simulations, general wave propagation simulations, comparisons with LIDAR data for positioning and / or other functions and / or other uses. In some embodiments, one or more tree traversal units (ttu) may be used to perform one or more ray tracing related operations.

[0204] One or more accelerators 614 (e.g., a cluster of hardware accelerators) have a wide range of uses in autonomous driving. PVA may be a programmable vision accelerator that can be used in key processing stages of ADA and autonomous vehicles. The capabilities of PVA are well suited for algorithmic domains that require predictable processing with low power consumption and low latency. In other words, PVA performs well on semi-intensive or regular intensive computations, even on small data sets, which require predictable runtimes with low latency and low power consumption. Therefore, in the context of autonomous vehicle platforms, PVA is designed to run classic computer vision algorithms because they are very efficient at object detection and integer math operations.

[0205] For example, according to one embodiment of the present technology, PVA is used to perform computer stereo vision. In some examples, a semi-global matching based algorithm can be used, although this is not intended to be limiting. Many applications of Level 3-5 autonomous driving require instant motion estimation / stereo matching (e.g., motion structure, pedestrian recognition, lane detection, etc.). PVA can perform computer stereo vision functions on input from two monocular cameras.

[0206] In some examples, PVA can be used to perform dense optical flow. Processed RADAR is provided based on processing raw RADAR data (e.g., using 4D fast Fourier transform). In other examples, PVA is used for time-of-flight depth processing, for example, by processing raw time-of-flight data to provide processed time-of-flight data.

[0207] DLA can be used to run any type of network to enhance control and driving safety, for example, including a neural network that outputs a confidence measure for each target detection. Such a confidence value can be interpreted as a probability, or provide a relative "weight" of each detection relative to other detections. This confidence value enables the system to further decide which detections should be considered true positive detections rather than false positive detections. For example, the system can set a threshold for confidence and only consider detections that exceed the threshold as true positive detections. In an automatic emergency braking (AEB) system, a false positive detection will cause the vehicle to automatically perform emergency braking, which is obviously undesirable. Therefore, only the most reliable detections should be considered as triggers for AEB. DLA can run a neural network to regress confidence values. The neural network can take as its input at least some subset of parameters, such as bounding box size, obtained ground plane estimation (e.g., from another subsystem), inertial measurement unit (IMU) sensor 666 output related to the vehicle 600 orientation and distance, and three-dimensional position estimation of objects obtained from neural networks and / or other sensors (e.g., LIDAR sensor 664 or RADAR sensor 660).

[0208] One or more SoCs 604 may include one or more data stores 616 (e.g., memory). Data store 616 may be on-chip memory of SoC 604 that may store neural networks to be executed on the GPU and / or DLA. In some examples, the capacity of data store 616 may be large enough to store multiple neural network instances for redundancy and safety. Data store 616 may include a L2 or L3 cache 612. As described herein, references to one or more data stores 616 may include references to memory associated with a PVA, DLA, and / or one or more other accelerators 614.

[0209] One or more SoCs 604 may include one or more processors 610 (e.g., embedded processors). The processor 610 may include a boot and power management processor, which may be a dedicated processor and subsystem for handling boot power and management functions and related security implementations. The boot and power management processor may be part of the boot sequence of one or more SoCs 604 and may provide runtime power management services. The boot power and management processor may provide clock and voltage programming, system low power state transition assistance, management of SoC 604 thermal sensors and temperature sensors, and / or management of SoC 604 power states. Each temperature sensor may be implemented as a ring oscillator whose output frequency is proportional to the temperature, and one or more SoCs 604 may use the ring oscillator to detect the temperature of one or more CPUs 606, one or more GPUs 608, and / or one or more accelerators 614. If it is determined that the temperature exceeds a threshold, the boot and power management processor may enter a temperature fault routine and place one or more SoCs 604 in a low power state and / or place the vehicle 600 in a driver safety parking mode (e.g., parking the vehicle 600 safely).

[0210] The one or more processors 610 may also include a set of embedded processors that can serve as an audio processing engine. The audio processing engine may be an audio subsystem that is capable of full hardware support for multi-channel audio through multiple interfaces and a wide and flexible audio I / O interface. In some examples, the audio processing engine is a dedicated processor core of a digital signal processor with dedicated RAM.

[0211] One or more processors 610 may also include an always-on processor engine that may provide the necessary hardware functionality to support low-power sensor management and wake-up use cases. The always-on processor engine may include a processor core, tightly coupled RAM, supporting peripherals (e.g., timers and interrupt controllers), various I / O controller peripherals, and routing logic.

[0212] One or more processors 610 may also include a safety cluster engine that includes a dedicated processor subsystem for handling safety management for automotive applications. The safety cluster engine may include two or more processor cores, tightly coupled RAM, supporting peripherals (e.g., timers, interrupt controllers, etc.), and / or routing logic. In safety mode, two or more cores may operate in lockstep mode and operate as a single core with comparison logic to detect any differences between their operations.

[0213] The one or more processors 610 may also include a real-time camera engine, which may include a dedicated processor subsystem for handling real-time camera management.

[0214] The one or more processors 610 may also include a high dynamic range signal processor, which may include an image signal processor, which is a hardware engine that is part of the camera processing pipeline.

[0215] One or more processors 610 may include a video image synthesizer, which may be a processing block (e.g., implemented on a microprocessor) that implements video post-processing functions required by the video playback application to generate the final image of the player window. The video image synthesizer may perform lens distortion correction on one or more wide-angle cameras 670, one or more surround cameras 674, and / or in-cabin monitoring camera sensors. The in-cabin monitoring camera sensors are preferably monitored by a neural network running on another instance of the advanced SoC that is configured to recognize in-cabin events and respond accordingly. The in-cabin system may perform lip reading to activate cellular service and make calls, dictate emails, change vehicle destinations, activate or change the vehicle's infotainment system and settings, or provide voice-activated web surfing. Certain features are only available to the driver when the vehicle is operating in automatic mode and are otherwise disabled.

[0216] The video image synthesizer may include enhanced temporal noise reduction for both spatial and temporal noise reduction. For example, in the case of motion in the video, the noise reduction appropriately weights the spatial information, thereby reducing the weight of information provided by adjacent frames. In the case where an image or portion of an image does not include motion, the temporal noise reduction performed by the video image synthesizer may use information from a previous image to reduce noise in the current image.

[0217] The video image compositor may also be configured to perform stereoscopic rectification on the input stereoscopic footage frames. The video image compositor may also be used for user interface composition when the operating system desktop is in use and does not require the GPU 608 to continuously render new surfaces. Even when one or more GPUs 608 are powered on and active in performing 3D rendering, the video image compositor may be used to offload one or more GPUs 608 to improve performance and responsiveness.

[0218] One or more SoCs 604 may also include a mobile industry processor interface (MIPI) camera serial interface for receiving video and input from a camera, a high-speed interface, and / or a video input block that may be used for a camera and related pixel input functions. One or more SoCs 604 may also include one or more input / output controllers that may be controlled by software and may be used to receive I / O signals that are not assigned a specific role.

[0219] One or more SoCs 604 may also include a wide range of peripheral interfaces to enable communication with peripherals, audio codecs, power management, and / or other devices. One or more SoCs 604 may be used to process data from cameras (e.g., via Gigabit multimedia serial links and Ethernet connections), sensors (e.g., one or more LIDAR sensors 664, one or more RADAR sensors 660, etc., which may be connected via Ethernet), data from bus 602 (e.g., speed of vehicle 600, steering wheel position, etc.), data from one or more GNSS sensors 658 (e.g., connected via Ethernet or CAN bus). One or more SoCs 604 may also include a dedicated high-performance mass storage controller, which may include its own DMA engine and may be used to free one or more CPUs 606 from routine data management tasks.

[0220] One or more SoC 604 can be an end-to-end platform with a flexible architecture that spans automation levels 3-5, thereby providing a comprehensive functional safety architecture that leverages and effectively uses computer vision and ADAS technologies to achieve diversity and redundancy, providing a platform for flexible and reliable driver software stacks and deep learning tools. Compared with traditional systems, one or more SoC 604 can be faster, more reliable, and even more energy-efficient and space-saving. For example, when one or more accelerators 614 are combined with one or more CPUs 606, one or more GPUs 608, and one or more data stores 616, a fast and efficient platform can be provided for level 3-5 autonomous vehicles.

[0221] As a result, the technology provides capabilities and functionality that are not possible with traditional systems. For example, computer vision algorithms can be executed on a CPU, which can be configured using a high-level programming language (such as the C programming language) to perform a wide variety of processing algorithms on a wide variety of visual data. However, CPUs often fail to meet the performance requirements of many computer vision applications, such as those related to execution time and power consumption. In particular, many CPUs are unable to execute complex object detection algorithms in real time, which is a requirement for in-vehicle ADAS applications and a requirement for practical Level 3-5 autonomous vehicles.

[0222] Compared to conventional systems, by providing a CPU complex, a GPU complex, and a hardware acceleration cluster, the techniques described herein allow for multiple neural networks to be executed simultaneously and / or sequentially, and for the results to be combined together to achieve Level 3-5 autonomous driving capabilities. For example, a CNN executed on a DLA or dGPU (e.g., one or more GPUs 620) may include text and word recognition, allowing the supercomputer to read and understand traffic signs, including signs for which the neural network has not been specifically trained. The DLA may also include a neural network capable of recognizing, interpreting, and providing semantic understanding of the symbols, and passing that semantic understanding to a path planning module running on the CPU complex.

[0223] As another example, multiple neural networks may be run simultaneously, such as may be required for Level 3, 4, or 5 driving. For example, a warning sign consisting of "WARNING: FLASHING LIGHTS INDICATE ICE CONDITIONS" and electric lights may be interpreted by multiple neural networks independently or in concert. The sign itself may be recognized as a traffic sign by a first deployed neural network (e.g., a trained neural network), and the text "FLASHING LIGHTS INDICATE ICE CONDITIONS" may be interpreted by a second deployed neural network, which, when detected, notifies the vehicle routing software (preferably executing on a CPU complex) that an icing condition exists. The flashing lights may be recognized by operating a third deployed neural network over multiple frames, notifying the vehicle routing software of the presence of the flashing lights. All three neural networks may be run simultaneously, such as within a DLA and / or on one or more GPUs 608.

[0224] In some examples, a CNN for facial recognition and owner recognition can use data from the camera sensor to identify the presence of an authorized driver and / or owner of the vehicle 600. The always-on sensor processing engine can be used to unlock the vehicle when the owner approaches the driver's door and turns on the lights, and, in a safety mode, disable the vehicle when the owner leaves the vehicle. In this way, one or more SoCs 604 provide anti-theft and / or carjacking protection.

[0225] In another example, a CNN for emergency vehicle detection and identification can use data from microphone 696 to detect and identify emergency vehicle sirens. Unlike conventional systems that use general classifiers to detect sirens and manually extract features, one or more SoC(s) 604 use CNNs to classify environmental and urban sounds, as well as classify visual data. In a preferred embodiment, the CNN running on the DLA is trained to identify the relative closing speeds of emergency vehicles (e.g., by using the Doppler effect). The CNN can also be trained to identify emergency vehicles specific to the area in which the vehicle is operating, as identified by one or more GNSS sensors 658. Thus, for example, when operating in Europe, the CNN will seek to detect European sirens, and when in the United States, the CNN will seek to identify only North American sirens. Once an emergency vehicle is detected, a control program can be used to execute an emergency vehicle safety routine with the help of ultrasonic sensors 662 to slow the vehicle, pull over, stop the vehicle, and / or idle the vehicle until one or more emergency vehicles pass.

[0226] The vehicle may include one or more CPUs 618 (e.g., one or more discrete CPUs or one or more dCPUs) that may be coupled to one or more SoCs 604 via a high-speed interconnect (e.g., PCIE). For example, one or more CPUs 618 may include an X86 processor. CPU 618 may be used to perform any of a variety of functions, including arbitrating potentially inconsistent results between ADAS sensors and SoC 604, and / or monitoring the status and health of one or more controllers 636 and / or infotainment SoC 630.

[0227] The vehicle 600 may include one or more GPUs 620 (e.g., one or more discrete GPUs or one or more dGPUs) that may be coupled to the SoC 604 via a high-speed interconnect (e.g., NVIDIA's NVLINK). The one or more GPUs 620 may provide additional artificial intelligence functionality, such as by executing redundant and / or different neural networks, and may be used to train and / or update the neural network based on input (e.g., sensor data) from sensors of the vehicle 600.

[0228] The vehicle 600 may also include a network interface 624, which may include one or more wireless antennas 626 (e.g., one or more wireless antennas for different communication protocols, such as a cellular antenna, a Bluetooth antenna, etc.). The network interface 624 may be used to enable wireless connections to the cloud (e.g., to one or more servers 678 and / or other network devices), to other vehicles, and / or to computing devices (e.g., a passenger's client device) via the Internet. In order to communicate with other vehicles, a direct link may be established between the two vehicles and / or an indirect link may be established (e.g., via a network and the Internet). A vehicle-to-vehicle communication link may be used to provide a direct link. The vehicle-to-vehicle communication link may provide the vehicle 600 with information about vehicles in the vicinity of the vehicle 600 (e.g., vehicles in front of, to the side of, and / or behind the vehicle 600). This functionality may be part of the cooperative adaptive cruise control functionality of the vehicle 600.

[0229] The network interface 624 may include a SoC that provides modulation and demodulation functions and enables one or more controllers 636 to communicate over a wireless network. The network interface 624 may include a radio frequency front end for up-conversion from baseband to radio frequency and down-conversion from radio frequency to baseband. Frequency conversion can be performed by well-known processes, and / or can be performed using a superheterodyne process. In some examples, the radio frequency front end function can be provided by a separate chip. The network interface may include a wireless function for communicating via LTE, WCDMA, UMTS, GSM, CDMA2000, Bluetooth, Bluetooth LE, Wi-Fi, Z-Wave, ZigBee, LoRaWAN, and / or other wireless protocols.

[0230] The vehicle 600 may further include one or more data stores 628, which may include off-chip (e.g., off-SoC). The data store 628 may include one or more storage elements, including RAM, SRAM, DRAM, VRAM, flash memory, hard disk, and / or other components and / or devices that can store at least one bit of data.

[0231] The vehicle 600 may also include one or more GNSS sensors 658. One or more GNSS sensors 658 (e.g., GPS, assisted GPS sensors, differential GPS (DGPS) sensors, etc.) are used to assist in mapping, perception, occupancy grid generation, and / or path planning functions. Any number of GNSS sensors 658 may be used, including, for example, but not limited to, GPS using a USB connector with an Ethernet to serial (RS-232) bridge.

[0232] The vehicle 600 may also include one or more RADAR sensors 660. Even in darkness and / or inclement weather conditions, the vehicle 600 may use one or more RADAR sensors 660 for remote vehicle detection. The RADAR functional safety level may be ASIL B. One or more RADAR sensors 660 may be controlled and access target tracking data using CAN and / or bus 602 (e.g., transmitting data generated by one or more RADAR sensors 660), and in some examples, access raw data by accessing Ethernet. A variety of RADAR sensor types may be used. For example, but not limited to, one or more RADAR sensors 660 may be suitable for front, rear, and side RADAR use. In some examples, a pulse Doppler RADAR sensor is used.

[0233] One or more RADAR sensors 660 may include different configurations, such as long-range with a narrow field of view, short-range with a wide field of view, short-range side coverage, etc. In some examples, the long-range RADAR can be used for adaptive cruise control functions. The long-range RADAR system can provide a wide field of view achieved by two or more independent scans, for example within a range of 250 meters. One or more RADAR sensors 660 can help distinguish between static and moving objects and can be used by the ADAS system for emergency braking assistance and forward collision warnings. The long-range RADAR sensor may include a single-base multimode RADAR with multiple (e.g., six or more) fixed RADAR antennas and high-speed CAN and FlexRay interfaces. In an example with six antennas, the central four antennas can create a focused beam pattern designed to record the 600 environment around the vehicle at a higher speed with minimal traffic interference in adjacent lanes. The other two antennas can expand the field of view, enabling it to quickly detect vehicles entering or leaving the vehicle's 600 lanes.

[0234] For example, a medium-range radar system may include a range of up to 160 meters (front) or 80 meters (rear) and a field of view of up to 42 degrees (front) or 150 degrees (rear). A short-range RADAR system may include, but is not limited to, a RADAR sensor designed to be mounted on either end of the rear bumper. When mounted on either end of the rear bumper, such a RADAR sensor system may produce two beams that continuously monitor the blind spots behind and to the sides of the vehicle.

[0235] Short-range RADAR systems may be used in ADAS systems for blind spot detection and / or lane change assistance.

[0236] The vehicle 600 may also include one or more ultrasonic sensors 662. One or more ultrasonic sensors 662 may be located at the front, rear, and / or sides of the vehicle 600 and may be used for parking assistance and / or creating and updating an occupancy grid. A variety of ultrasonic sensors 662 may be used, and different ultrasonic sensors 662 may be used for different detection ranges (e.g., 2.5 m, 4 m). One or more ultrasonic sensors 662 may operate at a functional safety level of ASIL B.

[0237] The vehicle 600 may include one or more LIDAR sensors 664. The one or more LIDAR sensors 664 may be used for object and pedestrian detection, emergency braking, collision avoidance, and / or other functions. The functional safety level of the one or more LIDAR sensors 664 may be ASIL B. In some examples, the vehicle 600 may include multiple LIDAR sensors 664 (e.g., two, four, six, etc.) that may use Ethernet (e.g., provide data to a Gigabit Ethernet switch).

[0238] In some examples, one or more LIDAR sensors 664 may be able to provide a list of objects and their distances in a 360-degree field of view. One or more commercial LIDAR sensors 664 have an advertised range of approximately 100m, an accuracy of 2cm-3cm, and support, for example, a 100Mbps Ethernet connection. In some examples, one or more non-protruding LIDAR sensors 664 may be used. In this example of the present invention, one or more LIDAR sensors 664 may be implemented as a small device that can be embedded in the front, rear, side, and / or corner of the vehicle 600. In this example of the present invention, one or more LIDAR sensors 664 may provide a horizontal field of view of up to 120 degrees and a vertical field of view of 35 degrees, with a range of 200m even for low reflectivity objects. The front-mounted one or more LIDAR sensors 664 may be configured with a horizontal field of view between 45 and 135 degrees.

[0239] In some examples, LIDAR technology, such as 3D flash LIDAR, can also be used. 3D flash LIDAR uses laser flash as a transmission source to illuminate about 200 meters around the vehicle. The flash LIDAR device includes a receiver that records the laser pulse transmission time and the reflected light on each pixel, and each pixel corresponds to the range from the vehicle to the object. Flash LIDAR can generate a highly accurate, distortion-free image of the environment with each laser flash. In some examples, four flash LIDAR sensors can be deployed, one on each side of the vehicle 600. Available 3D flash LIDAR systems include solid-state 3D staring array LIDAR cameras, with no moving parts except the fan (e.g., non-scanning LIDAR devices). The flash LIDAR device can use 5 nanosecond Class I (eye-safe) laser pulses per frame and can capture 3D distance point clouds and reflected lasers in the form of co-registered intensity data. By using flash LIDAR, and because flash LIDAR is a solid-state device with no moving parts, one or more LIDAR sensors 664 can be less susceptible to motion blur, vibration and / or shock.

[0240] The vehicle may also include one or more IMU sensors 666. In some examples, one or more IMU sensors 666 may be located at the center of the rear axle of the vehicle 600. One or more IMU sensors 666 may include, for example, but not limited to, one or more accelerometers, one or more magnetometers, one or more gyroscopes, one or more magnetic compasses, and / or other sensor types. In some examples, such as in a six-axis application, one or more IMU sensors 666 may include an accelerometer and a gyroscope, and in a nine-axis application, one or more IMU sensors 666 may include an accelerometer, a gyroscope, and a magnetometer.

[0241] In some embodiments, one or more IMU sensors 666 may be implemented as a miniature, high-performance GPS-aided inertial navigation system (GPS / INS) that combines micro-electromechanical systems (MEMS) inertial sensors, high-sensitivity GPS receivers, and advanced Kalman filter algorithms to provide estimates of position, velocity, and attitude. Thus, in some examples, one or more IMU sensors 666 may enable the vehicle 600 to estimate heading without input from a magnetic sensor by directly observing and correlating velocity changes from GPS to one or more IMU sensors 666. In some examples, one or more IMU sensors 666 and one or more GNSS sensors 658 may be combined in a single integrated unit.

[0242] The vehicle may include one or more microphones 696 positioned within and / or around the vehicle 600. The one or more microphones 696 may be used for emergency vehicle detection and identification, among other things.

[0243] The vehicle may also include any number of camera types, including one or more stereo cameras 668, one or more wide-angle cameras 670, one or more infrared cameras 672, one or more surround cameras 674, one or more long-range and / or mid-range cameras 698, and / or other camera types. The cameras may be used to capture image data of the entire periphery of the vehicle 600. The type of camera used depends on the embodiment and requirements of the vehicle 600, and any combination of camera types may be used to provide the necessary coverage around the vehicle 600. In addition, depending on the embodiment, the number of cameras may be different. For example, the vehicle may include six cameras, seven cameras, ten cameras, twelve cameras, and / or other numbers of cameras. As an example, the cameras may support, but are not limited to, Gigabit Multimedia Serial Link (GMSL) and / or Gigabit Ethernet. This document will refer to Fig. 6A and Figure 6B Describe each camera in more detail.

[0244] The vehicle 600 may also include one or more vibration sensors 642. One or more vibration sensors 642 may measure vibrations of vehicle components, such as axles. For example, changes in vibrations may indicate changes in the road surface. In another example, when two or more vibration sensors 642 are used, the difference between the vibrations may be used to determine friction or slippage of the road surface (e.g., when the vibration difference is between a powered drive shaft and a free-spinning shaft).

[0245] The vehicle 600 may include an ADAS system 638. In some examples, the ADAS system 638 may include a SoC. The ADAS system 638 may include automatic / adaptive / auto cruise control (ACC), cooperative adaptive cruise control (CACC), forward collision warning (FCW), automatic emergency braking (AEB), lane departure warning (LDW), lane keeping assist (LKA), blind spot warning (BSW), rear cross traffic warning (RCTW), collision warning system (CWS), lane centering (LC), and / or other features and functions.

[0246] The ACC system may use one or more RADAR sensors 660, one or more LIDAR sensors 664, and / or one or more cameras. The ACC system may include longitudinal ACC and / or lateral ACC. Longitudinal ACC monitors and controls the distance to the vehicle directly in front of the vehicle 600 and automatically adjusts the vehicle speed to maintain a safe distance from the vehicle in front. Lateral ACC performs distance keeping and suggests that the vehicle 600 change lanes when necessary. Lateral ACC is related to other ADAS applications, such as LCA and CWS.

[0247] CACC uses information from other vehicles, which can be received indirectly from other vehicles via a network interface 624 and / or one or more wireless antennas 626, via a wireless link, or via a network connection (e.g., via the Internet). A direct link can be provided by a vehicle-to-vehicle (V2V) communication link, while an indirect link can be an infrastructure-to-vehicle (I2V) communication link. In general, the V2V communication concept provides information about the vehicle in front (e.g., the vehicle directly in front of the vehicle 600 and in the same lane), while the I2V communication concept provides information about the traffic ahead. The CACC system may include one or both of the I2V and V2V information sources. Taking into account the vehicle information in front of the vehicle 600, CACC may be more reliable and has the potential to improve the smoothness of traffic flow and reduce congestion on the road.

[0248] The FCW system is designed to alert the driver to hazards so that the driver can take corrective action. The FCW system uses a forward-facing camera and / or one or more RADAR sensors 660 coupled to a dedicated processor, DSP, FPGA and / or ASIC, which is electrically coupled to driver feedback, such as a display, speaker and / or vibration component. The FCW system can provide warnings such as sound, visual warnings, vibrations and / or rapid brake pulses.

[0249] The AEB system detects an impending forward collision with another vehicle or other object, and if the driver does not take corrective action within specified time or distance parameters, the AEB system may automatically apply the brakes. The AEB system may use one or more front-facing cameras and / or one or more RADAR sensors 660 coupled to a dedicated processor, DSP, FPGA and / or ASIC. When the AEB system detects a hazard, it typically first alerts the driver to take corrective action to avoid the collision. If the driver does not take corrective action, the AEB system may automatically apply the brakes to prevent or at least mitigate the effects of the anticipated collision. The AEB system may include technologies such as dynamic brake support and / or collision approach braking.

[0250] The LDW system provides visual, audible, and / or tactile warnings, such as steering wheel or seat vibrations, to alert the driver when the vehicle 600 crosses a lane marking. The LDW system does not activate when the driver indicates an intention to leave the lane by activating a turn signal. The LDW system may use a front-facing camera coupled to a dedicated processor, DSP, FPGA, and / or ASIC that is electrically coupled to driver feedback, such as a display, speaker, and / or vibration assembly.

[0251] The LKA system is a variation of the LDW system. If the vehicle 600 begins to exit the lane, the LKA system provides steering input or braking to correct the vehicle 600. The BSW system detects and warns the driver of vehicles in the car's blind spot. The BSW system may provide visual, audible, and / or tactile alerts to indicate that it is unsafe to merge or change lanes. The system may provide additional warnings when the driver uses a turn signal. The BSW system may use a rear-facing camera and / or RADAR sensor.

[0252] The RCTW system may provide visual, audible, and / or tactile notifications when an object outside the range of the rear camera is detected while the vehicle 600 is in reverse. Some RCTW systems include AEB to ensure that the vehicle brakes are applied to avoid a collision. The RCTW system may use one or more rear-facing RADAR sensors 660 coupled to a dedicated processor, DSP, FPGA, and / or ASIC that is electrically coupled to driver feedback, such as a display, speaker, and / or vibration assembly.

[0253] Conventional ADAS systems may be prone to false positive results, which may be annoying and distracting to the driver, but generally do not have catastrophic consequences because the ADAS system alerts the driver and allows the driver to decide whether a safe condition actually exists and take appropriate action. However, in the autonomous vehicle 600, in the event of conflicting results, the vehicle 600 itself must decide whether to listen to the results from the main computer or the auxiliary computer (e.g., the first controller 636 or the second controller 636). For example, in some embodiments, the ADAS system 638 can be a backup and / or auxiliary computer for providing perception information to the backup computer module. The backup computer rationality monitor can run redundant different software on hardware components to detect failures in perception and dynamic driving tasks. The output from the ADAS system 638 can be provided to the monitoring MCU. If the outputs of the main computer and the auxiliary computer conflict, the monitoring MCU must determine how to reconcile the conflict to ensure safe operation.

[0254] In some examples, the primary computer may be configured to provide a confidence score to the monitoring MCU indicating the primary computer's confidence in the selected result. If the confidence score exceeds a threshold, the monitoring MCU may follow the direction of the primary computer regardless of whether the secondary computer provides conflicting or inconsistent results. If the confidence score does not meet the threshold and the primary and secondary computers indicate different results (e.g., conflict), the monitoring MCU may arbitrate between the computers to determine the appropriate result.

[0255] The monitoring MCU may be configured to run one or more neural networks that have been trained and configured to determine conditions under which the auxiliary computer provides a false alarm based on the outputs of the primary computer and the auxiliary computer. Thus, the one or more neural networks in the monitoring MCU may learn when the output of the auxiliary computer is credible and when it is not credible. For example, when the auxiliary computer is a radar-based FCW system, the one or more neural networks in the monitoring MCU may learn when the FCW system identifies a metal object that is not actually dangerous, such as a drain grate or manhole cover that triggers an alarm. Similarly, when the auxiliary computer is a camera-based LDW system, the neural network in the monitoring MCU may learn to override the LDW when a bicycle or pedestrian is present and a lane departure is actually the safest maneuver. In an embodiment including one or more neural networks running on the monitoring MCU, the monitoring MCU may include at least one of a DLA or a GPU suitable for running one or more neural networks with associated memory. In a preferred embodiment, the monitoring MCU may include and / or be included as a component of the SoC 604.

[0256] In other examples, the ADAS system 638 may include an auxiliary computer that uses traditional computer vision rules to perform ADAS functions. Thus, the auxiliary computer can use classic computer vision rules (if-then), and the presence of one or more neural networks in the monitoring MCU can improve reliability, safety, and performance. For example, diverse implementations and intentional non-identification make the entire system more fault-tolerant, especially to failures caused by software (or software-hardware interface) functions. For example, if there is a software defect or error in the software running on the main computer, and different software code running on the auxiliary computer provides the same overall result, the monitoring MCU can have greater confidence in the correctness of the overall result and that the software or hardware defect on the main computer did not cause a significant error.

[0257] In some examples, the output of the ADAS system 638 may be fed into the perception block of the host computer and / or the dynamic drive task block of the host computer. For example, if the ADAS system 638 indicates a forward collision warning due to an object directly ahead, the perception block may use that information in identifying the object. In other examples, as described herein, the secondary computer may have its own trained neural network, thereby reducing the risk of false positives.

[0258] The vehicle 600 may also include an infotainment SoC 630 (e.g., an in-vehicle infotainment system (IVI)). Although shown and described as an SoC, the infotainment system may not be a SoC and may include two or more discrete components. The infotainment SoC 630 may include a combination of hardware and software that can be used to provide audio (e.g., music, personal digital assistant, navigation instructions, news, radio, etc.), video (e.g., TV, movies, streaming, etc.), phone (e.g., hands-free calling), network connectivity (e.g., LTE, Wi-Fi, etc.), and / or provide information services to the vehicle 600 (e.g., navigation system, rear parking assist, radio data system, vehicle-related information such as fuel level, total distance traveled, brake fuel level, engine oil level, door open / closed, air filter information, etc.). For example, the infotainment SoC 630 may include a radio, a disk player, a navigation system, a video player, USB and Bluetooth connectivity, an onboard computer, in-car entertainment, Wi-Fi, steering wheel audio controls, hands-free voice control, a head-up display (HUD), an HMI display 634, a telematics device, a control panel (e.g., for controlling and / or interacting with various components, features, and / or systems), and / or other components. The infotainment SoC 630 may also be used to provide information (e.g., visual and / or auditory) to a vehicle user, such as information from an ADAS system 638, autonomous driving information, such as planned vehicle maneuvers, trajectories, surrounding environment information (e.g., intersection information, vehicle information, road information, etc.), and / or other information.

[0259] The infotainment SoC 630 may include GPU functionality. The infotainment SoC 630 may communicate with other devices, systems, and / or components of the vehicle 600 via a bus 602 (e.g., a CAN bus, Ethernet, etc.). In some examples, the infotainment SoC 630 may be coupled to a monitoring MCU so that the GPU of the infotainment system may perform some self-driving functions in the event of a failure of one or more main controllers 636 (e.g., a main computer and / or backup computer of the vehicle 600). In such an example, the infotainment SoC 630 may place the vehicle 600 in a driver-to-safety parking mode, as described herein.

[0260] The vehicle 600 may also include an instrument cluster 632 (e.g., a digital instrument panel, an electronic instrument cluster, a digital instrument cluster, etc.). The instrument cluster 632 may include a controller and / or a supercomputer (e.g., a discrete controller or a supercomputer). The instrument cluster 632 may include a set of instruments, such as a speedometer, a fuel level, an oil pressure, a tachometer, an odometer, a turn indicator, a shift position indicator, one or more seat belt warning lights, one or more parking brake warning lights, one or more engine fault lights, airbag (SRS) system information, lighting control, safety system control, navigation information, etc. In some examples, information may be displayed and / or shared between the infotainment SoC 630 and the instrument cluster 632. In other words, the instrument cluster 632 may be included as part of the infotainment SoC 630, and vice versa.

[0261] Fig.6D According to some embodiments of the present disclosure, one or more cloud-based servers and Fig. 6A 600. System diagram of communication between exemplary autonomous vehicles 600. System 676 may include one or more servers 678, one or more networks 690, and vehicles, including vehicle 600. One or more servers 678 may include multiple GPUs 684 (a) -684 (H) (collectively referred to as GPUs 684), PCIe switches 682 (a) -682 (H) (collectively referred to as PCIe switches 682), and / or CPUs 680 (A) -680 (B) (collectively referred to herein as CPUs 680). GPUs 684, CPUs 680, and PCIe switches may be interconnected via high-speed interconnects, such as, but not limited to, NVLink interfaces 688 and / or PCIE connections 686 developed by NVIDIA. In some examples, GPUs 684 are connected via NVLink and / or NV switch SoCs, and GPUs 684 and PCIe switches 682 are connected via PCIe interconnects. Although eight GPUs 684, two CPUs 680, and two PCIe switches are illustrated, this is not intended to be limiting. Depending on the embodiment, each of the one or more servers 678 may include any number of GPUs 684, CPUs 680, and / or PCIe switches. For example, one or more servers 678 may each include eight, sixteen, thirty-two, and / or more gpus 684.

[0262] One or more servers 678 may receive image data representing an image from a vehicle via one or more networks 690 that shows unexpected or changing road conditions, such as recently initiated road work. One or more servers 678 may send neural networks 692, updated neural networks 692, and / or map information 694 to the vehicle via one or more networks 690, including information about traffic and road conditions. Updates to map information 694 may include updates to HD map 622, such as information about construction sites, potholes, detours, flooding, and / or other obstacles. In some examples, neural network 692, updated neural network 692, and / or map information 694 may be derived from new training and / or experience represented in data received from any number of vehicles in the environment, and / or based on training performed at a data center (e.g., using one or more servers 678 and / or other servers).

[0263] One or more servers 678 can be used to train a machine learning model (e.g., a neural network) based on training data. The training data can be generated by the vehicle, and / or can be generated in simulation (e.g., using a game engine). In some examples, the training data is labeled (e.g., neural networks benefit from supervised learning) and / or undergoes other preprocessing, while in other examples, the training data is not labeled and / or preprocessed (e.g., neural networks do not require supervised learning). Training can be performed according to any one or more classes of machine learning techniques, including but not limited to: supervised training, semi-supervised training, unsupervised training, self-learning, reinforcement learning, joint learning, transfer learning, feature learning (including principal component and cluster analysis), multilinear subspace learning, manifold learning, representation learning (including alternate dictionary learning), rule-based machine learning, anomaly detection, and any variant or combination. Once the machine learning model is trained, the machine learning model can be used by the vehicle (e.g., transmitted to the vehicle 690 via one or more networks, and / or the machine learning model can be used by one or more servers 678 to remotely monitor the vehicle).

[0264] In some examples, one or more servers 678 may receive data from the vehicle and apply the data to the latest real-time neural network for real-time intelligent reasoning. One or more servers 678 may include a deep learning supercomputer and / or a dedicated AI computer powered by a GPU 684, such as the DGX and DGX Station machines developed by NVIDIA. However, in some examples, one or more servers 678 may include a deep learning infrastructure in a data center that uses only CPU power.

[0265] The deep learning infrastructure of one or more servers 678 can perform rapid real-time inference and can use this capability to assess and verify the health of the processor, software, and / or associated hardware in the vehicle 600. For example, the deep learning infrastructure can receive periodic updates from the vehicle 600, such as a sequence of images and / or objects that the vehicle 600 has located in the sequence of images (e.g., through computer vision and / or other machine learning object classification techniques). The deep learning infrastructure can run its own neural network to identify objects and compare them to the objects identified by the vehicle 600, and if the results do not match and the infrastructure concludes that the AI ​​in the vehicle 600 has failed, the one or more servers 678 can send a signal to the vehicle 600, instructing the fail-safe computer of the vehicle 600 to assume control, notify passengers, and complete a safe parking maneuver.

[0266] For inference, one or more servers 678 may include one or more GPUs 684 and one or more programmable inference accelerators (e.g., NVIDIA's TensorRT). The combination of GPU-powered servers and inference acceleration may enable real-time responses. In other examples, such as where performance is less critical, servers powered by CPUs, FPGAs, and other processors may be used for inference.

[0267] Example computing device

[0268] Figure 7 700 is a block diagram of an example computing device 700 suitable for implementing some embodiments of the present disclosure. The computing device 700 may include an interconnect system 702 that directly or indirectly couples the following devices: a memory 704, one or more central processing units (CPUs) 706, one or more graphics processing units (GPUs) 708, a communication interface 710, an input / output (I / O) port 712, an input / output component 714, a power supply 716, one or more presentation components 718 (e.g., a display), and one or more logic units 720. In at least one embodiment, the one or more computing devices 700 may include one or more virtual machines (VMs), and / or any of their components may include virtual components (e.g., virtual hardware components). For non-limiting examples, the one or more GPUs 708 may include one or more vGPUs, the one or more CPUs 706 may include one or more vCPUs, and / or the one or more logic units 720 may include one or more virtual logic units. Thus, one or more computing devices 700 may include discrete components (eg, a complete GPU dedicated to computing device 700 ), virtual components (eg, a portion of a GPU dedicated to computing device 700 ), or a combination thereof.

[0269] although Figure 7The various modules in are shown as being connected with lines through the interconnection system 702, but this is not intended to be limiting and is only for clarity. For example, in some embodiments, a presentation component 718 such as a display device can be considered an I / O component 714 (e.g., if the display is a touch screen). As another example, the CPU 706 and / or the GPU 708 can include a memory (e.g., the memory 704 can also represent a storage device in addition to the memory of the GPU 708, the CPU 706, and / or other components). In other words, Figure 7 The computing devices in the description are illustrative only. No distinction is made between "workstations," "servers," "laptops," "desktops," "tablets," "client devices," "mobile devices," "handheld devices," "game consoles," "electronic control units (ECUs)," "virtual reality systems," and / or other device or system types, as in Figure 7 as contemplated within the scope of computing devices.

[0270] The interconnection system 702 may represent one or more links or buses, such as an address bus, a data bus, a control bus, or a combination thereof. The interconnection system 702 may include one or more bus or link types, such as an industry standard architecture (ISA) bus, an extended industry standard architecture (EISA) bus, a video electronics standard association (VESA) bus, a peripheral component interconnect (PCI) bus, a peripheral component interconnect pass-through (PCIE) bus, and / or other types of buses or links. In some embodiments, there is a direct connection between components. For example, the CPU 706 may be directly connected to the memory 704. In addition, the CPU 706 may be directly connected to the GPU 708. In the case where there is a direct connection or a point-to-point connection between components, the interconnection system 702 may include a PCIe link for performing the connection. In these examples, the PCI bus does not need to be included in the computing device 700.

[0271] Memory 704 may include any of a variety of computer-readable media. Computer-readable media may be any available media that can be accessed by computing device 700. Computer-readable media may include volatile and non-volatile media and removable and non-removable media. By way of example and not limitation, computer-readable media may include computer storage media and communication media.

[0272] Computer storage media may include volatile and nonvolatile media and / or removable and non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, program modules, and / or other data types. For example, memory 704 may store computer-readable instructions (e.g., instructions representing one or more programs and / or one or more program elements), such as an operating system. Computer storage media may include, but are not limited to, RAM, ROM, EEPROM, flash memory or other storage technology, CD-ROM, digital versatile disk (DVD) or other optical disk storage, cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by computing device 700. As used herein, computer storage media itself does not include signals.

[0273] Computer storage media may carry computer readable instructions, data structures, program modules, and / or other data types contained in a modulated data signal (e.g., a carrier wave or other transport mechanism), and include any information delivery media. The term "modulated data signal" may refer to a signal that has one or more of its characteristics set or changed in a manner that encodes the information in the signal. By way of example and not limitation, computer storage media may include wired media, such as a wired network or direct wired connection, and wireless media, such as acoustic, RF, infrared, and other wireless media. Any combination of the above should also be included within the scope of computer readable media.

[0274] One or more CPUs 706 may be configured to execute at least some computer-readable instructions to control one or more components of the computing device 700 to perform one or more methods and / or processes described herein. One or more CPUs 706 may each include one or more cores (e.g., one, two, four, eight, twenty-eight, seventy-two, etc.) capable of processing multiple software threads simultaneously. One or more CPUs 706 may include any type of processor, and may include different types of processors (e.g., processors with fewer cores for mobile devices and processors with more cores for servers) according to the type of computing device 700 implemented. For example, depending on the type of computing device 700, the processor may be an Advanced RISC Machine (ARM) processor implemented using Reduced Instruction Set Computing (RISC) or an x86 processor implemented using Complex Instruction Set Computing (CISC). In addition to one or more microprocessors or supplementary coprocessors (e.g., math coprocessors), the computing device 700 may also include one or more cpu 706.

[0275] In addition to one or more CPUs 706, or selected from one or more CPUs 706, one or more GPUs 708 can be configured to execute at least some computer-readable instructions to control one or more components of the computing device 700 to perform one or more methods and / or processes described herein. One or more GPUs 708 can be integrated GPUs (e.g., with one or more CPUs 706 and / or one or more GPUs 708 can be discrete GPUs). In an embodiment, one or more GPUs 708 can be coprocessors of one or more CPUs 706. The computing device 700 can use one or more GPUs 708 to render graphics (e.g., 3D graphics) or perform general-purpose computing. For example, one or more GPUs 708 can be used for general-purpose computing on a GPU (GPGPU). One or more GPUs 708 can include hundreds or thousands of cores that can process hundreds or thousands of software threads simultaneously. One or more GPUs 708 may generate pixel data for outputting an image in response to a rendering command (e.g., a rendering command received from one or more CPUs 706 via a host interface). One or more GPUs 708 may include graphics memory, such as display memory, for storing pixel data or any other suitable data, such as GPGPU data. The display memory may be included as part of the memory 704. One or more GPUs 708 may include two or more GPUs operating in parallel (e.g., via a link). The link may connect the GPUs directly (e.g., using NVLINK) or connect the GPUs through a switch (e.g., using NVSwitch). When combined together, each GPU 708 may generate pixel data or GPGPU data for different parts of the output or different outputs (e.g., a first GPU for a first image and a second GPU for a second image). Each GPU may include its own memory or may share memory with other GPUs.

[0276] In addition to or selected from one or more CPUs 706 and / or one or more GPUs 708, one or more logic units 720 may be configured to execute at least some computer-readable instructions to control one or more components of the computing device 700 to perform one or more methods and / or processes described herein. In an embodiment, one or more CPUs 706, one or more GPUs 708, and / or one or more logic units 720 may discretely or jointly perform any combination of methods, processes, and / or portions thereof. One or more of the logic units 720 may be a part of and / or integrated into one or more of the CPUs 706 and / or GPUs 708 and / or one or more of the logic units 720 may be a discrete component or otherwise located outside of one or more CPUs 706 and / or GPUs 708, and one or more logic units 720 may be a coprocessor of one or more CPUs 706 and / or GPUs 708.

[0277] Examples of one or more logic units 720 include one or more processing cores and / or components thereof, such as a data processing unit (DPU), a tensor core (TC), a tensor processing unit (TPU), a pixel vision core (PVC), a vision processing unit (VPU), a graphics processing cluster (GPC), a texture processing cluster (TPC), a streaming multiprocessor (SM), a tree traversal unit (TTU), an artificial intelligence accelerator (AIA), a deep learning accelerator (DLA), an arithmetic logic unit (ALU), an application specific integrated circuit (ASIC), a floating point unit (FPU), an input / output (I / O) element, a peripheral component interconnect (PCI) or a peripheral component interconnect express (PCIe) element, and / or the like.

[0278] The communication interface 710 may include one or more receivers, transmitters, and / or transceivers that enable the computing device 700 to communicate with other computing devices via an electronic communication network including wired and / or wireless communications. The communication interface 710 may include components and functions to support communication via any of a number of different networks, such as wireless networks (e.g., Wi-Fi, Z-Wave, Bluetooth, Bluetooth LE, ZigBee, etc.), wired networks (e.g., communicating via Ethernet or InfiniBand), low-power wide area networks (e.g., LoRaWAN, SigFox, etc.), and / or the Internet. In one or more embodiments, the one or more logic units 720 and / or the communication interface 710 may include one or more data processing units (DPUs) for transmitting data received via a network and / or via the interconnect system 702 directly to one or more GPUs 708 (e.g., their memory).

[0279] The I / O ports 712 may enable the computing device 700 to be logically coupled to other devices, including I / O components 714, one or more presentation components 718, and / or other components, some of which may be built-in (e.g., integrated into) the computing device 700. Illustrative I / O components 714 include microphones, mice, keyboards, joysticks, game pads, game controllers, satellite dishes, scanners, printers, wireless devices, and the like. The I / O components 714 may provide a natural user interface (NUI) that processes air gestures, voice, or other physiological inputs generated by a user. In some cases, the input may be transmitted to an appropriate network element for further processing. The NUI may implement any combination of voice recognition, stylus recognition, facial recognition, biometrics, gesture recognition on and near the screen, air gestures, head and eye tracking, and touch recognition associated with the display of the computing device 700 (as described in more detail in the present disclosure). The computing device 700 may include a depth camera, such as a stereo camera system, an infrared camera system, an RGB camera system, touch screen technology, and combinations of these, for gesture detection and recognition. Additionally, computing device 700 may include an accelerometer or gyroscope capable of detecting motion (e.g., as part of an inertial measurement unit (IMU)). In some examples, the output of the accelerometer or gyroscope may be used by computing device 700 to render immersive augmented reality or virtual reality.

[0280] The power supply 716 may include a hardwired power supply, a battery power supply, or a combination thereof. The power supply 716 may provide power to the computing device 700 to enable the components of the computing device 700 to operate.

[0281] The one or more presentation components 718 may include a display (e.g., a display, a touch screen, a television screen, a head-up display (HUD), other display types, or combinations thereof), speakers, and / or other presentation components. The one or more presentation components 718 may receive data from other components (e.g., one or more GPUs 708, one or more CPUs 706, etc.) and output data (e.g., as images, video, sound, etc.).

[0282] Sample Data Center

[0283] Figure 8 An example data center 800 is shown that can be used in at least one embodiment of the present disclosure. The data center 800 can include a data center infrastructure layer 810, a framework layer 820, a software layer 830, and / or an application layer 840.

[0284] like Figure 8 As shown, the data center infrastructure layer 810 may include a resource coordinator 812, group computing resources 814, and node computing resources ("node CRs") 816 (1)-816 (N), where "N" represents any integer. In at least one embodiment, the node CRs 816 (1)-816 (N) may include, but are not limited to, any number of central processing units (CPUs) or other processors (including DPUs, accelerators, field programmable gate arrays (FPGAs), graphics processors or graphics processing units (GPUs), etc.), storage devices (e.g., dynamic read-only memories), in some embodiments, storage devices (e.g., solid-state or disk drives), network input / output (NW I / O) devices, network switches, virtual machines (VMs), power modules and / or cooling modules, etc., and one or more of the node CRs 816 (1)-816 (N) may correspond to a server having one or more of the above computing resources. Furthermore, in some embodiments, node CR816(1)-816(N) may include one or more virtual components, such as a vGPU, a vCPU, and / or the like, and / or one or more of the node CR816(1)-816(N) may correspond to a virtual machine (VM).

[0285] In at least one embodiment, the grouping computing resources 814 may include a separate grouping of node CR816 located in one or more racks (not shown), or multiple racks located in a data center at different geographical locations (also not shown). The separate grouping of node CR816 in the grouping computing resources 814 may include grouping computing, network, memory or storage resources, which may be configured or allocated to support one or more workloads. In at least one embodiment, multiple node CR816 including CPU, GPU, DPU and / or other processors may be grouped in one or more racks to provide computing resources to support one or more workloads. One or more racks may also include any number of power modules, cooling modules and / or network switches in any combination.

[0286] Resource coordinator 812 may configure or otherwise control one or more node CRs 816(1)-816(N) and / or grouped computing resources 814. In at least one embodiment, resource coordinator 812 may comprise a software design infrastructure (SDI) management entity for data center 800. Resource coordinator 812 may comprise hardware, software, or some combination thereof.

[0287] In at least one embodiment, Figure 8 As shown, the framework layer 820 may include a job scheduler 833, a configuration manager 834, a resource manager 836, and / or a distributed file system 838. The framework layer 820 may include a framework that supports software 832 of the software layer 830 and / or one or more applications 842 of the application layer 840. The software 832 or one or more applications 842 may include web-based service software or applications, such as services provided by Amazon Web Services, Google Cloud, and Microsoft Azure. The framework layer 820 may be, but is not limited to, a free open source software web application framework, such as Apache SparkTM (hereinafter referred to as "Spark"), which can utilize the distributed file system 838 for large-scale data processing (e.g., "big data"). In at least one embodiment, the job scheduler 833 may include a Spark driver to facilitate the scheduling of workloads supported by various layers of the data center 800. The configuration manager 834 may configure different layers, such as the software layer 830 and the framework layer 820, including Spark and the distributed file system 838, to support large-scale data processing. The resource manager 836 may be capable of managing cluster or group computing resources mapped to or allocated to support the distributed file system 838 and the job scheduler 833. In at least one embodiment, the cluster or group computing resources may include group computing resources 814 at the data center infrastructure layer 810. The resource manager 836 may coordinate with the resource coordinator 812 to manage these mapped or allocated computing resources.

[0288] In at least one embodiment, software 832 included in software layer 830 may include software used by node CRs 816(1)-816(N), group computing resources 814, and / or at least a portion of a distributed file system 838 of framework layer 820. The one or more software may include, but are not limited to, Internet web page search software, email virus scanning software, database software, and streaming video content software.

[0289] In at least one embodiment, the applications 842 included in the application layer 840 may include one or more types of applications used by the node CRs 816(1)-816(N), the grouped computing resources 814, and / or at least a portion of the distributed file system 838 of the framework layer 820, but are not limited to any number of genomics applications, perceptual computing, and machine learning applications, including training or inference software, machine learning framework software (e.g., PyTorch, TensorFlow, Caffe, etc.), and / or other machine learning applications used in conjunction with one or more embodiments.

[0290] In at least one embodiment, any of the configuration manager 834, resource manager 836, and resource coordinator 812 can implement any number and type of self-modification actions based on any number and type of data acquired in any technically feasible manner. The self-modification actions can save the data center operator of the data center 800 from making potentially erroneous configuration decisions and potentially avoid underutilized and / or underperforming portions of the data center.

[0291] According to one or more embodiments described herein, data center 800 may include tools, services, software, or other resources for training one or more machine learning models or predicting or inferring information using one or more machine learning models. For example, one or more machine learning models may be trained by calculating weight parameters according to a neural network architecture using software and / or computing resources described in this disclosure with respect to data center 800. In at least one embodiment, a trained or deployed machine learning model corresponding to one or more neural networks may be used to infer or predict information using resources described in this disclosure with respect to data center 800 using weight parameters calculated using one or more training techniques (such as, but not limited to, the training techniques described herein).

[0292] In at least one embodiment, the data center 800 may use a CPU, an application-specific integrated circuit (ASIC), a GPU, an FPGA, and / or other hardware (or corresponding virtual computing resources) to perform training and / or inference using the above resources. In addition, one or more software and / or hardware resources described in the present disclosure may be configured as a service to allow a user to train or perform information reasoning, such as image recognition, speech recognition, or other artificial intelligence services.

[0293] Example network environment

[0294] A network environment suitable for implementing embodiments of the present disclosure may include one or more client devices, servers, network attached storage (NAS), other backend devices, and / or other device types. The client devices, servers, and / or other device types (e.g., each device) may be configured to communicate with the client device, server, or other device types. Figure 7 The data center 800 may be implemented on one or more instances of one or more computing devices 700 - for example, each device may include similar components, features and / or functions of the computing device 700. In addition, backend devices (servers, NAS, etc.) may be included as part of the data center 800, examples of which are referred to herein. Figure 8 Describe in more detail.

[0295] The components of the network environment can communicate with each other through one or more networks that are wired, wireless, or both. The network may include multiple networks, or a network of networks. For example, the network may include one or more wide area networks (WANs), one or more local area networks (LANs), one or more public networks (e.g., the Internet and / or a public switched telephone network (PSTN)), and / or one or more private networks. In the case where the network includes a wireless telecommunications network, components such as base stations, communication towers, or even access points (as well as other components) may provide wireless connections.

[0296] Compatible network environments may include one or more peer-to-peer network environments, in which case the server may not be included in the network environment, and one or more client-server network environments, in which case one or more servers may be included in the network environment. In a peer-to-peer network environment, the functionality described herein with respect to the server may be implemented on any number of client devices.

[0297] In at least one embodiment, the network environment may include one or more cloud-based network environments, distributed computing environments, combinations thereof, and the like. A cloud-based network environment may include a framework layer, a job scheduler, a resource manager, and a distributed file system implemented on one or more servers, which may include one or more core network servers and / or edge servers. The framework layer may include a framework for software supporting the software layer and / or one or more applications of the application layer. The software or one or more applications may include web-based service software or applications, respectively. In an embodiment, one or more client devices may use web-based service software or applications (e.g., by accessing the service software and / or applications via one or more application programming interfaces (APIs)). The framework layer may be, but is not limited to, a free, open source software web application framework, for example, a distributed file system may be used for large-scale data processing (e.g., "big data").

[0298] A cloud-based network environment may provide cloud computing and / or cloud storage that performs any combination of the computing and / or data storage functions (or one or more portions thereof) described herein. Any of these various functions may be distributed to multiple locations from a central or core server (e.g., a server in one or more data centers), which may be distributed in a state, a region, a country, the world, etc. If the connection to a user (e.g., a client device) is relatively close to one or more edge servers, one or more core servers may assign at least a portion of the functions to one or more edge servers. A cloud-based network environment may be private (e.g., limited to a single organization), public (e.g., available to many organizations), and / or a combination thereof (e.g., a hybrid cloud environment).

[0299] One or more client devices may include the Figure 7 At least some of the components, features, and functionality of one or more of the example computing devices 700 are described. By way of example and not limitation, a client device may be embodied as a personal computer (PC), a laptop computer, a mobile device, a smartphone, a tablet computer, a smart watch, a wearable computer, a personal digital assistant (PDA), an MP3 player, a virtual reality headset, a global positioning system (GPS) or device, a video player, a camera, a surveillance device or system, a vehicle, a boat, an aircraft, a virtual machine, a drone, a robot, a handheld communication device, a hospital device, a gaming device or system, an entertainment system, a vehicle computer system, an embedded system controller, a remote control, an appliance, a consumer electronic device, a workstation, an edge device, any combination of the depicted devices, or any other suitable device.

[0300] The present disclosure may be described in the general context of computer code or machine usable instructions, including computer executable instructions, such as program modules, executed by a computer or other machine (such as a personal data assistant or other handheld device). Typically, program modules include routines, programs, objects, components, data structures, etc., which refer to code that performs a specific task or implements a specific abstract data type. The present disclosure may be implemented in a variety of system configurations, including handheld devices, consumer electronics, general-purpose computers, more specialized computing devices, etc. The present disclosure may also be implemented in a distributed computing environment, where tasks are performed by remote processing devices linked through a communication network.

[0301] As used herein, the description of "and / or" about two or more elements should be interpreted as referring to only one element or a combination of elements. For example, "element A, element B, and / or element C" may include only element A, element B, element C, element A and element B, element A and element C, element B and element C, or element A, element B, and element C. In addition, "at least one of element A or element B" may include at least one of element A, at least one of element B, or at least one of element A and at least one of element B. In addition, "at least one of element A and element B" may include at least one of element A, at least one of element B, or at least one of element A and at least one of element B. In addition, the use of the term "based on" should not be interpreted as "based only on" or "only based on." Instead, a first element "based on" a second element includes instances where the first element is based on the second element but may also be based on one or more additional elements.

[0302] In order to satisfy statutory requirements, the subject matter of the present disclosure is described in detail herein. However, the description itself is not intended to limit the scope of the present disclosure. On the contrary, the inventors have contemplated that the claimed subject matter may also be embodied in other ways, including different steps or combinations of steps similar to the steps described in this document, as well as other existing or future technologies. In addition, although the terms "step" and / or "block" may be used herein to imply different elements of the method employed, the terms should not be interpreted as implying any particular order between the various steps disclosed herein unless the order of the various steps is explicitly described.

Claims

1. A system comprising: One or more processing units, the one or more processing units configured to perform operations, the operations comprising: Obtaining execution information corresponding to a first task and a second task, wherein the first task includes one or more first operations, and the second task includes one or more second operations, the execution information indicating: a first dependency corresponding to the first task; a second dependency corresponding to the second task; a first computing platform on which the first task is to be performed; a second computing platform different from the first computing platform on which the second task is to be executed; and current availability of computing resources corresponding to the first computing platform and the second computing platform; determining a time to initialize execution of the first task and the second task based at least on the execution information; and Execution of the first task on the first computing platform and the second task on the second computing platform is initialized at the time, and at least one first operation of the one or more first operations and at least one second operation of the one or more second operations are interdependent.

2. The system according to claim 1, wherein the execution information further indicates: a first timeout period corresponding to the first task, indicating a time period after which execution of the one or more first operations is terminated; and A second timeout period corresponding to the second task indicates a time period after which execution of the one or more second operations is terminated.

3. The system of claim 2, wherein the first timeout period and the second timeout period are the same.

4. The system of claim 2, wherein the operations further comprise: Based at least on reaching the first timeout period and the second timeout period, execution of the first task on the first computing platform and the second task on the second computing platform is cancelled.

5. The system of claim 1, wherein the operations further comprise: detecting one or more errors during the execution of the first task or during the execution of the second task; as well as Based at least on the detected one or more errors, execution of the first task or execution of the second task is terminated.

6. The system of claim 1, wherein the first computing platform and the second computing platform comprise one or more of a CPU, a GPU, a TPU, an FPGA, a DPU, a PPU, an ASIC, or a DLA.

7. The system of claim 1, wherein the system is included in at least one of the following: control systems for autonomous or semi-autonomous machines; Perception systems for autonomous or semi-autonomous machines; A system for performing simulation operations; Systems for performing digital twin operations; A system for performing light transport simulations; A system for performing collaborative content creation of 3D assets; Systems for performing deep learning operations; a system for presenting at least one of augmented reality content, virtual reality content, or mixed reality content; A system for hosting one or more real-time streaming applications; Systems implemented using edge devices; Systems implemented using robots; Systems for performing conversational AI operations; A system implementing one or more large language models (LLMs); Systems for generating synthetic data; A system comprising one or more virtual machines VM; A system for performing one or more generative AI operations; A system implemented at least in part in a data center; or A system implemented at least in part using cloud computing resources.

8. A system comprising: One or more processing units, the one or more processing units configured to perform operations, the operations comprising: Obtaining execution information corresponding to a plurality of tasks, the plurality of tasks respectively including one or more containers, the one or more containers respectively including one or more organized operations, the execution information indicating dependencies respectively corresponding to respective tasks of the plurality of tasks, computing resources on which respective operations corresponding to respective tasks of the plurality of tasks are executed, and current availability of the computing resources; determining a plurality of stages for initiating execution of operations corresponding to the plurality of tasks, the plurality of stages being sequentially ordered based at least on the execution information; and Initializing execution of the plurality of tasks, wherein: A first task in the plurality of tasks is performed using a computing platform of a first type, and a second task in the plurality of tasks is performed using a computing platform of a second type; The execution of one or more first containers corresponding to the first task and the execution of one or more second containers corresponding to the second task are interdependent; and The first task and the second task are executed in the same phase of the plurality of phases.

9. The system according to claim 8, wherein the execution information further indicates: a first timeout period corresponding to the first task, indicating a time period after which execution of the one or more first containers is terminated; and A second timeout period corresponding to the second task indicates a time period after which execution of the one or more second containers is terminated.

10. The system of claim 9, wherein the first timeout period and the second timeout period are the same.

11. The system of claim 9, the operations further comprising: Based at least on reaching the first timeout period and the second timeout period, execution of the first task on the first type of computing platform and the second task on the second type of computing platform is cancelled.

12. The system of claim 8, the operations further comprising: detecting one or more errors during the execution of the first task or during the execution of the second task; as well as Based at least on the detected one or more errors, execution of the first task or execution of the second task is terminated.

13. A system according to claim 8, wherein the execution information also includes one or more priority levels assigned to one or more tasks among the plurality of tasks, a higher priority level assigned to a particular task indicating that the particular task is executed before one or more other tasks assigned with a lower priority level.

14. The system of claim 7, wherein the system is included in at least one of the following: control systems for autonomous or semi-autonomous machines; Perception systems for autonomous or semi-autonomous machines; A system for performing simulation operations; Systems for performing digital twin operations; A system for performing light transport simulations; A system for performing collaborative content creation of 3D assets; Systems for performing deep learning operations; a system for presenting at least one of augmented reality content, virtual reality content, or mixed reality content; A system for hosting one or more real-time streaming applications; Systems implemented using edge devices; Systems implemented using robots; Systems for performing conversational AI operations; A system for performing one or more generative AI operations; A system implementing one or more large language models (LLMs); Systems for generating synthetic data; A system comprising one or more virtual machines VM; A system implemented at least in part in a data center; or A system implemented at least in part using cloud computing resources.

15. A method comprising: Obtaining execution information corresponding to a first task and a second task, the first task comprising one or more first containers, the one or more first containers comprising one or more first operations, the second task comprising one or more second containers, the one or more second containers comprising one or more second operations, the execution information indicating a first dependency corresponding to the first task, a second dependency corresponding to the second task, a first computing platform on which the first task is to be executed, a second computing platform different from the first computing platform on which the second task is to be executed, and current availability of computing resources corresponding to the first computing platform and the second computing platform; as well as Initializing execution of the first task on the first computing platform and the second task on the second computing platform such that, based at least on the execution information, the one or more first containers are interdependent with the one or more second containers.

16. The method according to claim 15, wherein the execution information further comprises: a first timeout period corresponding to the first task, indicating a time period after which execution of the one or more first containers is terminated; as well as A second timeout period corresponding to the second task indicates a time period after which execution of the one or more second containers is terminated. The method of claim 16 , wherein the first timeout period and the second timeout period are the same.

18. The method according to claim 16, further comprising: Based at least on reaching the first timeout period and the second timeout period, execution of the first task on the first computing platform and the second task on the second computing platform is cancelled.

19. The method according to claim 15, further comprising: detecting one or more errors during the execution of the first task or during the execution of the second task; as well as Based on the detected one or more errors, execution of the first task or execution of the second task is terminated.

20. The method of claim 15, wherein the first computing platform and the second computing platform comprise one or more of a CPU, a GPU, a TPU, an FPGA, a DPU, a PPU, an ASIC, or a DLA.

Images